The Role of FOXM1 in Breast Cancer Chemotherapy Resistance

The role of FOXM1 in breast cancer chemotherapy resistance

Thesis submitted by

Pasarat Khongkow Department of Surgery and Cancer, Division of Cancer, Imperial College London, Hammersmith Hospital Campus

London, United Kingdom

Imperial College London

For the degree of Doctor of Philosophy

2015

DECLARATION OF ORIGINALITY

Unless otherwise stated in the text, the work presented in this thesis, including the experiments, analyses and discussions, is entirely the result of my own work.

The immunohistochemistry studies in chapter 3 and 4 were performed and analysed with the collaborators: Prof. Ui-Soon Khoo, Ms. Chun Cong, and Ms. Ellen P. S. Man from Department of Pathology, the University of Hong Kong.

‘The copyright of this thesis rests with the author and is made available under a

Creative Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work’

ABSTRACT

Resistance to chemotherapeutic agents is the main obstacle to the effective breast cancer management. Therefore, it is important to elucidate the molecular mechanisms of chemoresistance and develop novel therapeutic strategies in order to overcome drug resistance. In this work, I found that FOXM1 is a critical mediator of epirubicin and paclitaxel resistance in MCF-7 breast cancer cell lines. FOXM1 expression was upregulated in both epirubicin resistant MCF-7 (MCF-7 EpiR) and paclitaxel resistant MCF-7 (MCF-7 TaxR) cells compared to sensitive MCF-7 cells.

Interestingly, its depletion dramatically impaired the clonogenic survival and significantly induced cellular senescence in the resistant cells. In addition, I identified two novel downstream FOXM1 targets, NBS1 and KIF20A, involved in epirubicin and paclitaxel resistance, respectively.

Firstly, I found that FOXM1 transcriptionally regulated NBS1 expression to modulate

HR-mediated DSB repair and epirubicin resistance. Overexpression of FOXM1 and

NBS1 lead to the enhancement of HR efficiency to eliminate epirubicin-induced DNA damage. Conversely, similar to FOXM1, depletion of NBS1 also sensitised both

MCF-7 and MCF-7 EpiR cells to epirubicin by inducing cellular senescence.

Secondly, I identified the mitotic kinesin KIF20A as a direct downstream target of

FOXM1, participating in the mitotic spindle formation and paclitaxel resistance.

Depletion of KIF20A caused mitotic spindle abnormalities, inhibition of cell growth as well as the induction of senescent cells in both MCF-7 and MCF-7 TaxR cells.

Consistently, immunohistochemical analysis of breast cancer patient samples revealed that high expression levels of FOXM1, NBS1 and KIF20A are strongly correlated with poor prognosis in breast cancer, supporting a physiological role of

FOXM1 and its novel targets in genotoxic drug resistance.

Collectively, these findings suggests that FOXM1 and its targets, NBS1 and KIF20A, could be reliable prognostic markers for monitoring treatment efficiency as well as promising targets for therapeutic intervention to overcome epirubicin and paclitaxel resistance in breast cancer.

PUBLICATIONS

Parts of this work have been published in:

1. KHONGKOW, P., GOMES, A. R., GONG, C., MAN, E. P. S., TSANG, J. W. H., ZHAO, F., MONTEIRO, L. J., COOMBES, R. C., MEDEMA, R. H., KHOO, U. S. & LAM, E. W. F. 2015. Paclitaxel targets FOXM1 to regulate KIF20A in mitotic catastrophe and breast cancer paclitaxel resistance. Oncogene.

2. KHONGKOW, P., KARUNARATHNA, U., KHONGKOW, M., GONG, C., GOMES, A. R., YAGUE, E., MONTEIRO, L. J., KONGSEMA, M., ZONA, S., MAN, E. P. S., TSANG, J. W. H., COOMBES, R. C., WU, K. J., KHOO, U. S., MEDEMA, R. H., FREIRE, R. & LAM, E. W. F. 2014. FOXM1 targets NBS1 to regulate DNA damage-induced senescence and epirubicin resistance. Oncogene, 33, 4144-4155.

3. ZHAO, F., SIU, M. K. Y., JIANG, L., TAM, K. F., NGAN, H. Y. S., LE, X. F., WONG, O. G. W., WONG, E. S. Y., GOMES, A. R., BELLA, L., KHONGKOW, P., LAM, E. W. F. & CHEUNG, A. N. Y. 2014. Overexpression of Forkhead Box Protein M1 (FOXM1) in Ovarian Cancer Correlates with Poor Patient Survival and Contributes to Paclitaxel Resistance. PLoS ONE, 9, e113478.

4. MYATT, S. S., KONGSEMA, M., MAN, C. W. Y., KELLY, D. J., GOMES, A. R., KHONGKOW, P., KARUNARATHNA, U., ZONA, S., LANGER, J. K., DUNSBY, C. W., COOMBES, R. C., FRENCH, P. M., BROSENS, J. J. & LAM, E. W. F. 2014. SUMOylation inhibits FOXM1 activity and delays mitotic transition. Oncogene, 33, 4316-4329.

5. KHONGKOW, P., KARUNARATHNA, U. & LAM, E. 2013. Abstract A92: The role of FOXM1 and NBS1 in DNA damage-induced senescence and epirubicin resistance. Molecular Cancer Therapeutics, 12, A92.

6. KHONGKOW, M., OLMOS, Y., GONG, C., GOMES, A. R., MONTEIRO, L. J., YAGÜE, E., CAVACO, T. B., KHONGKOW, P., MAN, E. P. S., LAOHASINNARONG, S., KOO, C.-Y., HARADA-SHOJI, N., TSANG, J. W.-H., COOMBES, R. C., SCHWER, B., KHOO, U.-S. & LAM, E. W.-F. 2013. SIRT6 modulates paclitaxel and epirubicin resistance and survival in breast cancer. Carcinogenesis, 34, 1476-1486.

7. MONTEIRO, L. J., KHONGKOW, P., KONGSEMA, M., MORRIS, J. R., MAN, C., WEEKES, D., KOO, C. Y., GOMES, A. R., PINTO, P. H., VARGHESE, V., KENNY, L. M., CHARLES COOMBES, R., FREIRE, R., MEDEMA, R. H. & LAM, E. W. F. 2012. The Forkhead Box M1 protein regulates BRIP1 expression and DNA damage repair in epirubicin treatment. Oncogene

8. KHONGKOW, P., KARUNARATHNA, H. U. & LAM, E. W. 2012 The Role of FOXM1 and NBS1 in DNA Double Strand Breaks Repair and Epirubicin Resistance. European Journal of Cancer, 48, S173, 728.

9. KHONGKOW, P, MIDDLETON, A., WONG, J., KANDOLA, N., KONGSEMA, M., NESTAL DE MORAES, G., GOMES, A. R. & LAM, E. W. F. In vitro methods for studying the mechanisms of resistance to DNA damaging therapeutic drugs. Methods in Molecular Biology: Cancer Drug Resistance. Springer. Book chapter in preparation.

To my beloved family Thank you for everything

ACKNOWLEDGEMENT

Firstly, I would like to express my deepest gratitude to my supervisor Professor Eric

Lam for his attention, guidance, and patience throughout my PhD study. His enthusiasm and inspiring discussions kept me going. Thank you, Eric, this thesis would not have been possible without your guidance, encouragement and constant support.

Secondly, I would like to thank Dr Ana Gomes for her kind help and endless support throughout my PhD study. Additionally, I would like to thank Upekha Karunaratha for her invaluable friendship and for staying by my side in every moments of weakness or happiness.

Thanks to all previous and present colleagues in Lam’s lab for their contributions to the scientific and social environment. I would like to extend my gratitude to our collaborators: Prof. Ui-Soon Khoo, Ms. Chan Cong, and Dr. Ellen P. S. Man from

Department of Pathology, the University of Hong Kong for their help in immunohistochemistry studies. I would like to thank my friends along these years,

Sasiwan Laohasinnaroung, Mesayamas Khongsema, and Chonlathep Usaku for their patience, support and friendship.

My special thanks go to my beloved fiancé, Amnart Kassanook, thank you for your unconditional love, patience, and support throughout this long journey. Without you, I would not have completed this thesis. Five years being far from you was tough, but now it’s time to start the next chapter of our life, together.

My deepest appreciation goes to Mattaka Khongkow, my beloved twin sister, my little angel, my best friend, my sunshine, and my everything. You don’t know how much you mean to me. Thank you for everything, I could have never gotten to this point

8 without you. I would also like to thank my little sister, Jarupa Khongkow, thank you for your love and support. Love you both.

I would like to express my heartfelt thanks to my beloved parents, Suthep and

Ratana Khongkow, for your never-ending love and support. I am extremely lucky to be your daughter. Especially mom, you are my real inspiration to do research about breast cancer with the hope that one day we will beat cancer. I would not be the person I am today without you both, love you. I also dedicate this thesis to my beloved grandparents, Luan and Cha-on Khongkow, who have always been proud of me. Five years were too long for you to wait, but I hope that you could see me graduate and move on to the next stage in my life, I miss you.

Finally, thank you the Royal Thai Government Scholarship for providing me a great opportunity to fulfil my dream here at Imperial College. I will do my best to be a good researcher for our country.

‘ขอบคุณทุกก ำลังใจ ขอบคุณทุกมิตรภำพ ขอบคุณทุกควำมช่วยเหลือที่ท ำให้มีวันนี้’

Pasarat Khongkow

TABLE OF CONTENTS

DECLARATION OF ORIGINALITY ...... 2

ABSTRACT...... 3

PUBLICATIONS ...... 5

ACKNOWLEDGEMENT...... 8

TABLE OF CONTENTS ...... 10

TABLE OF FIGURES ...... 15

LIST OF TABLES...... 19

LIST OF SUPPLEMENTARY FIGURES ...... 20

ABBREVIATIONS ...... 21

CHAPTER 1 INTRODUCTION...... 28

1.1 Breast cancer ...... 29

1.1.1 Risk factors ...... 29

1.1.2 Medical management of breast cancer ...... 30

1.1.3 Molecular classifications of breast cancer ...... 31

1.2 Chemotherapy ...... 33

1.2.1 Anthracyclines ...... 33

1.2.2 Paclitaxel ...... 36

1.3 Apoptosis and non-apoptotic deaths in treatment response ...... 37

1.3.1 Cellular senescence ...... 39

1.3.2 Mitotic Catastrophe ...... 42

1.4 General mechanisms of drug resistance in breast cancer ...... 45

1.5 DNA Damage Response (DDR) ...... 48

1.5.1 DNA double strand break repair ...... 51

1.5.2 NBS1 ...... 55

1.6 Cell cycle and cancer ...... 57

1.6.1 Cell cycle checkpoints ...... 58

1.7 Kinesins and Cancer ...... 63

1.7.1 KIF20A ...... 64

1.8 FOXM1 ...... 66

1.8.1 FOXM1 gene and its protein structure ...... 66

1.8.2 FOXM1 in cell cycle and proliferation ...... 68

1.8.3 FOXM1 in tumorigenesis and cancer progression ...... 71

1.8.4 FOXM1 and DNA damage ...... 73

1.8.5 FOXM1 and Chemotherapy resistance ...... 74

1.9 Thesis Aims ...... 76

CHAPTER 2 MATERIALS AND METHODS...... 77

2.1 Cell culture and cell lines ...... 78

2.1.1 The human breast cancer cell lines: MCF-7 and MDA-MB-231 ...... 78

2.1.2 MCF-7 EpiR and MCF-7 TaxR ...... 78

2.1.3 Mouse embryonic fibroblasts (MEFs) ...... 78

2.1.3 HeLa cells carrying DR-GFP...... 79

2.1.4 Maintaining Cultured Cells ...... 79

2.1.5 Cell line storage ...... 79

2.2 Drug treatment ...... 80

2.3 Plasmid constructs ...... 80

2.3.1 Cloning of WT and mut KIF20A luciferase reporter constructs ...... 81

2.4 Transfection ...... 81

2.4.1 Gene Silencing with Small Interfering RNAs (siRNAs) ...... 81

2.5 Quantitative real time PCR ...... 85

2.5.1 RNA extraction and quantification ...... 85

2.5.2 cDNA synthesis (Reverse transcription) ...... 85

2.5.3 Primers Design and optimisation ...... 86

2.5.4 Real time quantitative PCR (RT-qPCR) ...... 87

2.6 Western Blot Analysis ...... 88

2.6.1 Protein extraction and quantification ...... 88

2.6.2 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)89

2.6.4 Antibodies ...... 90

2.7 Immunofluorescent Staining ...... 92

2.7.1 γ-H2AX assay ...... 92

2.7.2 α-tubulin immunostaining to detect mitotic catastrophe ...... 93

2.8 Luciferase reporter assay ...... 94

2.9 HR Repair Assay ...... 94

2.10 Sulforhodamine B (SRB) assay ...... 95

2.11 Clonogenic Assay ...... 96

2.12 Senescence-Associated β-Galactosidase assay ...... 97

2.13 Cell Cycle Analysis ...... 97

2.14 Chromatin Immunoprecipitation (ChIP) ...... 98

2.15 Clinical data analysis ...... 100

2.15.1 Tissue Microarray ...... 100

2.15.2 Immnohistochemistry ...... 100

2.15.3 Staining scoring and Statistical Analysis ...... 101

2.16 Statistic Analysis ...... 102

CHAPTER 3 THE ROLE OF FOXM1 IN DNA DAMAGE RESPONSE AND EPIRUBICIN RESISTANCE ...... 103

3.1 Introduction ...... 104

3.2 Results ...... 106

3.2.1 FOXM1-deficient MEFs exhibited increased levels of DNA damages upon epirubicin treatment ...... 106

3.2.2 FOXM1 reconstitution in Foxm1-/- MEFs exhibits the decreased accumulation of γH2AX foci in response to epirubicin treatment ...... 107

3.2.3 FOXM1 deletion significantly inhibits long-term clonogenic ability and induces premature senescence in response to DNA damage...... 111

3.2.4 Knockdown of FOXM1 leads to increased levels of DNA damage in MCF-7 EpiR cells...... 114

3.2.5 Knockdown of FOXM1 significantly sensitises MCF-7 breast cancer cells to epirubicin-induced cellular senescence ...... 116

3.2.6 FOXM1 and DNA repair gene, NBS1, are up-regulated in MCF-7 EpiR cells ...... 119

3.2.7 FOXM1 enhances NBS1 expression and ATM activity to promote DNA damage repair signalling ...... 121

3.2.8 FOXM1-deficient cells display decreased expression of the DNA repair gene NBS1 ...... 123

3.2.9 Overexpression of FOXM1 leads to upregulation of NBS1 expression and enhances ATM activity...... 125

3.2.10 FOXM1 regulates NBS1 expression through a FHRE in its promoter .. 126

3.2.11 FOXM1 and NBS1 required for efficient HR-mediated repair ...... 127

3.2.12 Ectopic expression of NBS1 in Foxm1-/- MEFs reduces the sensitivity to epirubicin, as indicated by the decreased accumulation of γH2AX foci ...... 129

3.2.13 Depletion of NBS1 increases sensitivity to epirubicin and induces senescence associated phenotypes in MCF-7 breast cancer cells ...... 131

3.2.14 Overexpression of FOXM1 or NBS1 confers resistance to epirubicin possibly by enhancing DNA repair pathway in MCF-7 cells ...... 136

3.2.15 Correlation between NBS1 and FOXM1 expression in breast cancer samples ...... 138

3.3 Discussion ...... 141

CHAPTER 4 THE ROLE OF FOXM1 IN MITOTIC CONTROL AND PACLITAXEL RESISTANCE ...... 147

4.1 Introduction ...... 148

4.2 Results ...... 150

4.2.1 Deletion of FOXM1 reduces long-term clonogenic survival and induces cellular senescence in response to paclitaxel treatment ...... 150

4.2.2 Knockdown of FOXM1 suppresses cell proliferation in both paclitaxel sensitive and resistance MCF-7 cells ...... 152

4.2.3 Elevated expression of FOXM1 confers paclitaxel resistance...... 154

4.2.4 KIF20A and FOXM1 are upregulated in paclitaxel resistant MCF-7 (MCF-7 TaxR) cells ...... 155

4.2.5 Silencing of FOXM1 downregulates KIF20A expression ...... 157

4.2.6 Overexpression of FOXM1 enhances the transcriptional activity of KIF20A promoter ...... 160

4.2.7 FOXM1 directly binds to KIF20A loci in MCF-7 cells...... 161

4.2.8 Low concentrations of paclitaxel cause aberrant mitosis in MCF-7 cells 164

4.2.9 Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation and chromosome alignment...... 167

4.2.10 Intracellular localisation of FOXM1 and KIF20A during mitosis ...... 172

4.2.11 Depletion of KIF20A or FOXM1 reduces long-term clonogenic survival and induces cellular senescence in MCF-7 and MCF-7 TaxR cell lines...... 174

4.2.12 Correlation between FOXM1 and KIF20A expression in breast cancer samples ...... 179

4.3 Discussion ...... 185

CHAPTER 5 FINAL DISCUSSION ...... 190

SUPPLEMENTARY DATA ...... 208

REFERENCES ...... 214

TABLE OF FIGURES

Figure 1.1 Structures of the anthracyclines, doxorubicin and epirubicin……………34

Figure 1.2 Molecular structure of Paclitaxel…………………………………………. 36

Figure 1.3 Different chemotherapy-induced cell death mechanisms……………… 38

Figure 1.4 Molecular mechanisms of anti-cancer drug resistance…………………. 47

Figure 1.5 The model of the ATM signalling pathway in response to DSBs……….50

Figure 1.6 DNA double strand break repair mechanisms…………………………… 54

Figure 1.7 The structure and protein interaction sites of human NBS1/Nibrin……. 56

Figure 1.8 Four main stages of mitosis……………………………………………….. 58

Figure 1.9 The FOXM1 gene and its mRNA and protein structures……………….. 67

Figure 1.10 FOXM1 is a crucial cell cycle regulator…………………………………. 70

Figure 1.11 Functional roles of FOXM1………………………………………………..75

Figure 3.1 Foxm1-/- MEFs exhibit higher levels of DNA breaks than WT MEFs after epirubicin treatment ……………………………………………………………………...108

Figure 3.2 Foxm1-/- MEFs expressing ectopic FOXM1-mcherry plasmids exhibit decreased DNA breaks after epirubicin treatment……………………………………110

Figure 3.3 FOXM1 deletion inhibits clonogenic growth and induces cellular senescence in response to DNA damage in MEFs……………………...... 113

Figure 3.4 FOXM1 depletion leads to increased levels of DNA damage…...... 115

Figure 3.5 Knockdown of FOXM1 suppresses cell growth and induces cellular senescence in MCF-7 and MCF-7 EpiR cells………………………………………….117

Figure 3.6 Knockdown of FOXM1 causes the accumulation of γH2AX foci in response to γ-irradiation…………………………………………………………………118

Figure 3.7 Epirubicin resistant cell line exhibits increased FOXM1 and NBS1 protein and mRNA expression levels……………………………………………………………120

Figure 3.8 MCF-7 and MCF-7 EpiR cells depleted in FOXM1 expression reveal significant decreases in both protein and mRNA levels of NBS1…………………...122

Figure 3.9 FoxM1-deficient MEFs exhibited decreased mRNA and protein levels of the DNA repair gene NBS1 compared with wild type MEFs…………………………124

Figure 3.10 FOXM1 regulates NBS1 expression and modulates ATM phosphorylation…………………………………………………………………………...125

Figure 3.11 FOXM1 regulates NBS1 expression through a FHRE site within its promoter…………………………………………………………………………………...126

Figure 3.12 FOXM1 and NBS1 are required for efficient HR repair……………..…128

Figure 3.13 Ectopic expression of NBS1 in Foxm1-/- MEFs reduces the accumulation of γH2AX foci……………………………………………………………………………...130

Figure 3.14 The depletion of NBS1 increases cell sensitivity to epirubicin……..…132

Figure 3.15 MCF-7 EpiR cells depleted in NBS1 expression reveal an increase in expression of cleaved PARP protein…………………………………………………...133

Figure 3.16 NBS1 depletion induces senescence associated-phenotypes in MC

F-7 breast cancer cells…………………………………………………………………..134

Figure 3.17 Overexpression of FOXM1 or NBS1 promotes epirubicin resistance and decrease DNA damage in MCF-7 cells………………………………………………..137

Figure 3.18 Correlation between NBS1 and FOXM1 expression in breast cancer samples……………………………………………………………………………………139

Figure 3.19 Kaplan-Meier analysis of overall survival for FOXM1 and NBS1 mRNA transcript expression……………………………………………………………………..140

Figure 3.20 A novel possible role of FOXM1 in DNA damage response pathway.144

Figure 4.1 Deletion of FOXM1 reduces long-term clonogenic survival and induces cellular senescence in response to paclitaxel treatment in MEFs…………………..152

Figure 4.2 FOXM1 knockdown suppresses the clonogenic ability of MCF-7 and MCF-7 TaxR cells…………………………………………………………………………154

Figure 4.3 Overexpression of FOXM1 induces resistance to paclitaxel in MCF-7 cells………………………………………………………………………………………...155

Figure 4.4 FOXM1 and KIF20A expression levels are strongly upregulated in paclitaxel-resistant MCF-7 cells……….………………………………………………..157

Figure 4.5 Silencing of FOXM1 downregulates KIF20A expression in MCF-7 and MCF-7 TaxR cells…………………………………………………………………………159

Figure 4.6 Depletion of FOXM1 causes the decreased expression of KIF20A in MDA-MB-231 cells………………………………………………………………………..160

Figure 4.7 Overexpression of FOXM1 enhances the transcriptional activity of KIF20A promoter in MCF-7 cells………………………………………………………..164

Figure 4.8 FOXM1 directly binds to the KIF20A promoter region in MCF-7 cells..165

Figure 4.9 Low concentrations of paclitaxel cause mitotic catastrophe…………...167

Figure 4.10 Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation and chromosome alignment…………………………………………………170

Figure 4.11 Depletion of FOXM1 or KIF20A causes anaphase bridges, lagging chromosomes, and chromosome instability in MCF-7 cells…………………………172

Figure 4.12 Characterisation of mitotic spindle defects in MCF-7 cells following by FOXM1 or KIF20A knockdown………………………………………………………….173

Figure 4.13 Dynamic localisation of FOXM1 and KIF20A during mitosis of MCF-7 cells…………………………………………………………………………...……………175

Figure 4.14 The transfection efficiency of FOXM1-siRNA and KIF20A-siRNA in MCF-7 cells………………………………………………………………………………..177

Figure 4.15 Targeting KIF20A or FOXM1 using siRNA inhibits cell growth and induces senescence in MCF-7 cells…………………………………...……………….178

Figure 4.16 The transfection efficiency of FOXM1-siRNA and KIF20A-siRNA in MCF-7 EpiR cells……………...…………………………………………………………..179

Figure 4.17 Targeting KIF20A or FOXM1 using siRNA significantly impairs colony formation and induces senescence in MCF-7 TaxR cells…………………………….180

Figure 4.18 KIF20A overexpression is significantly associated with poorer survival in breast cancer patients……………………………………………..…………………….183

Figure 4.19 Upregulation of KIF20A is significantly associated with poorer survival in breast cancer patients…………………………………………………….……………..184

Figure 4.20 Univariate and multivariate Cox-regression analysis using KIF20A nuclear score and other clinicopathological parameters………………………….….185

Figure 5.1 Schematic representation summarising the possible role of FOXM1 in the development of epirubicin resistance……………………………………………..……209

Figure 5.2 Schematic representation summarising the possible role of FOXM1 in the development of paclitaxel resistance…………………………………………………..210

LIST OF TABLES

Table 1.1 Major molecular subtypes of breast cancer determined by gene expression profiling…………………………………………………………..………….…32

Table 1.2 Cell death pathway characteristics……….………………………………….44

Table 2.1 List of human RT-qPCR primers used in this study………………………..86

Table 2.2 List of mouse RT-qPCR primers used in this study…………………..……86

Table 2.3 Components of the difference percentage of SDS-PAGE gels………...…89

Table2.4 List of primary antibodies used for western blotting………...………………91

Table 2.5 Primers used in ChIPs………………...………………………………………99

Table 2.6 List of antibodies used for immunohistochemistry……………..…………101

LIST OF SUPPLEMENTARY FIGURES

Figure S1 Quantification of γH2AX foci using FociCounter software………………212

Figure S2 FOXM1 depletion leads to dramatically increases levels of DNA damage in MCF-7 cells………………………………………………………………………….....213

Figure S3 Optimization of epirubicin treatment doses used in long-term clonogenic assay for WT and Foxm1−/− MEFs………………………………………...…………...214

Figure S4 Densitometric analysis of the western blots in figure 3.7A……...………215

Figure S5 Densitometric analysis of the western blots in figure 3.8…………..……216

Figure S6 Densitometric analysis of the western blots in figure 4.4A………..….…217

Figure S7 Densitometric analysis of the western blots in figure 4.5B…………...…218

Figure S8 Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation in MCF-7 TaxR cells……………………………………………………….….210

Figure S9 FOXM1 was unable to transactivate a putative KIF20A promoter region containing the 5’ UTR (−1150/−61) in the MCF-7 cells…….… ..……………………220

Figure S10 Cloning of WT, MUT1 and MUT2 KIF20A luciferase reporter constructs………………………………………………………………..………………..221

Figure S11 Depletion of FOXM1 increases p21 expression and induces G/M arrest in both MCF-7 and MCF-7 TaxR cells…………..………………………………………222

ABBREVIATIONS

AML Acute myeloid leukaemia

APC/C Anaphase-Promoting Complex

APS Ammonium persulphate

ATF2 Activating transcription factor 2

ATM Ataxia telangiectasia mutated

ATR Ataxia telangiectasia and Rad3 related

BCL-2 B-cell lymphoma 2

BCRP Breast cancer resistance protein

Bmi-1 B cell-specific Moloney murine leukemia virus integration site 1

BRAF v-Raf murine sarcoma viral oncogene homolog B

BRCA1 Breast cancer type 1 susceptibility protein

BRCA2 Breast cancer type 2 susceptibility protein

BRCT Breast cancer C-terminus

BRIP1 BRCA1 interacting protein C-terminal helicase 1

BSA Bovine serum albumin

TBS-T Tris-buffered saline and tween

Bub3 Budding uninhibited by benzimidazoles 3 protein

CDC20 Cell division cycle 20 protein

CDC25A Cell division cycle 25 homolog A

CDC25C Cell division cycle 25 homolog C

CDH1 Cadherin-1

CDK1 Cyclin B1-dependent kinase

PLK Polo-like kinases cDNA Complementary DNA

CDH1 Cadherin 1

CENP-A Centromere protein A

CENP-F Centromere protein F

ChIP Chromation-Immunoprecipitation

CHK1 Checkpoint kinase 1

CHK2 Checkpoint kinase 2

CKS1 Cyclin-dependent kinases regulatory subunit 1

CtBP C-terminal-binding protein 1

CtIP CtBP-interacting protein

DAPI 4',6-diamidino-2-phenylindole

DDR DNA damage response

DEPC Diethylpyrocarbonate

DMSO Di-methyl sulfoxide

DNA2 DNA replication helicase/nuclease 2

DNA-PK DNA-dependent protein kinase

DNA-PKcs DNA-dependent protein kinase catalytic subunit

DOX Doxorubicin

DR-GFP Direct repeat-green flouorescent protein

DSBs DNA double-strand breaks

DTT Dithiothreitol

EDTA Ethylenediaminetetraacetic acid

EGTA Ethylene glycol tetraacetic acid

EMT Epithelial-mesenchymal transition

EPI Epirubicin

ER Estrogen receptor

EV Empty vector control

EXO1 Exonuclease 1

FCS Foetal calf serum

FHA Forkhead-associated domain

FKH Forkhead DNA Binding Domain

FOXM1 Forkhead box protein M1

FOXO3a Forkhead box O3

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HER2 Human epidermal growth factor receptor 2

HFH-11 Hepatocyte nuclear factor 3/fork head homolog 11

HJ Holiday junction

HNSCC Head and neck squamous cell carcinoma

HR Homologous recombination

HRP Horseradish peroxidase

HRT Hormone replacement therapy

IgG Immunoglobulin G

INCENP Inner centromere protein

IR Ionising radiation

JNK1 c-Jun N-terminal kinases

KIF Kinesin

KIF20A Kinesin family member 20A

LOX Lysyl oxidase

Mad2 Mitotic arrest deficient 2

MAPs Microtubule associated proteins

MDM2 Mouse double minute 2 homolog

MDR Multidrug resistance

MEK MAPK/ERK kinase mESC Mouse embryonic stem cells miRNA/ miR microRNA

MKLP2 Mitotic kinesin-like protein 2

MMP-2 Matrix metalloproteinase-2

MMP-9 Matrix metalloproteinase-9

MPP2 Membrane protein, palmitoylated 2

MRE11 Meiotic recombination 11

MRN MRE11-RAD50-NBS1 complex

MRP-1 Multidrug resistance protein -1

MTs Microtubules

Mut Mutant

Na3VO4 Sodium orthovanadate

NaF Sodium fluoride

NBS Nijmegen breakage syndrome

NBS1 Nijmegen breakage syndrome gene

NHEJ Non-homologous end joining

NRD N-terminal Repressor Domain

NS siRNA Non-silencing control siRNA

OIS Oncogene-induced senescence

OSCC Oral squamous cell carcinoma

PALB2 Partner and localizer of BRCA2

PBS Phosphate buffered saline

PCR Polymerase chain reaction

PDAC Pancreatic ductal adenocarcinoma

PgP P-glycoprotein

PI Protease inhibitor

PI-3K Phosphoinositide 3-kinase

PMSF Phenylmethysulfonyl fluoride

PR Progesterone receptor

PTEN Phosphatase and tensin homolog p16 p16INK4A or cyclin-dependent kinase inhibitor 2A p21 p21 CIP1/WAF1 or cyclin-dependent kinase inhibitor 1

RAB6KIFL Rab6-binding kinesin

RAD50 DNA repair protein RAD50

RB Retinoblastoma protein

RIPA Radioimmunoprecipitation assay buffer

RNF20 Ring finger protein 20

RPA Replication protein A

RR Relative risk

RS Replicative senescence

RT-qPCR Real-time quantitative reverse transcription PCR

S.D. Standard deviation

SAC Spindle assembly checkpoint

SAHF Senescence-associated heterochromatic foci

SASP Senescence-associated secretory phenotype

SA-βgal Senescence-associated β galactosidase

SDS Sodium Dodecyl Sulphate Reagent

SDS-PAGE Sulfate-polyacrylamide gel electrophoresis siRNA Small interfering RNA

SIRT1 Sirtuin 1

SKP2 S-phase kinase-associated protein 2

SMC1 Structural maintenance of chromosomes protein 1

SRB Sulforhodamine B ssDNA Single-stranded DNA

STK11 Serine/threonine-protein kinase 11

SUMO Small ubiquitin-related modifier

TAD Transactivation Domain

Tax Paclitaxel

TCA Trichloroacetic acid

TE Tris-EDTA buffer solution

TEMED Tetraethylethyleneduanine

TIS Therapy-induced senescence

TMA Tissue microarray

TOP2 Topoisomerase II

TOPBP1 DNA topoisomerase 2-binding protein 1

TP53 Tumour protein p53

VEGF Vascular endothelial growth factor

WIN Winged helix

WRN Werner syndrome ATP-dependent helicase

WT Wild-type wtNIP Wild-type NBS1 inhibitory peptides

XLF XRCC4-like factor

XRCC4 X-ray repair cross-complementing protein 4

γH2AX Phosphorylation of the histone H2AX

CHAPTER 1

INTRODUCTION

1.1 Breast cancer

Globally, breast cancer is the most frequently diagnosed malignancy among women.

There are about 1.38 million new cases and 458,000 deaths per year from breast cancer, making it the most common cause of women’s cancer deaths worldwide

(Ferlay et al., 2010). In the UK, breast cancer remains the most common cancer in women and its incidence continues to increase (CRUK, 2015). However, breast cancer mortality rates have fallen by over 25% in the past two decades; this is largely the result of early detection as well as substantial improvements in breast cancer management (Turner and Jones, 2008, Yeo et al., 2014). It is estimated that approximately one in nine women who survive to the age of 85 will be diagnosed with breast cancer during their lifetime (Kelsey and Berkowitz, 1988).

1.1.1 Risk factors

The causes of breast cancer are complex. It has been suggested that breast cancer involves the combined effects of numerous genetic, environmental, and behavioural risk factors that are unique to each individual. Age is one of the most important risk factors for breast cancer. The chances of developing the disease increase with age.

About 95% of women diagnosed with breast cancer each year are over age 40, and about half are aged 61 and older (CRUK, 2015). Reproductive factors are the other main breast cancer risk factors, including early onset of menstruation, late age of first pregnancy, fewer pregnancies, shorter or no periods of breastfeeding, and a later menopause. It therefore seems that the most established breast cancer risk factors are thought to influence risk through hormone-related pathways, higher concentrations of endogenous oestrogens are strongly associated with increased risk for breast cancer in postmenopausal women, and trials have shown that the anti- oestrogen tamoxifen reduces the incidence of breast cancer (Travis and Key, 2003).

The long-term use of exogenous sex hormones, including oestrogen and progestogen, may also increase the risk of breast cancer development, such as oral contraceptive (Beral et al., 1996) or hormone replacement therapy (HRT) (Million

Women Study, 2002). In addition, the burden of breast cancer is also associated with non-reproductive lifestyle factors including the increase in obesity and alcohol consumption, as well as a lack of physical activity (Howell et al., 2014). A strong family history with breast cancer is another well-known risk factor, particularly when the first degree relatives have been diagnosed with the disease. However, only 5-

10% of the all breast cancer cases are thought to be hereditary, meaning that they are a direct result of germline mutations in highly penetrable genes inherited from a parent, such as Breast cancer susceptibility gene 1(BRCA1) and Breast cancer susceptibility gene 2 (BRCA2) (Wooster et al., 1995, Da Silva and Lakhani, 2010,

Miki et al., 1994). Recently, various other genes, including Checkpoint kinase 2

(CHK2), Phosphatase and tensin homolog (PTEN), Tumor protein p53 (TP53),

Ataxia telangiectasia mutated (ATM), Serine/threonine-protein kinase 11 (STK11),

Cadherin 1 (CDH1), Nijmegen breakage syndrome (NBS1), DNA repair protein

RAD50 (RAD50), BRCA1-interacting protein 1 (BRIP1) and Partner and localizer of

BRCA2 (PALB2) have been identified to confer an increased risk of hereditary breast cancer (van der Groep et al., 2011). Numerous epidemiological risk factors have also been identified to increase a woman’s risk of developing breast cancer, but the cause of any individual breast cancer is mostly unknown.

1.1.2 Medical management of breast cancer

Over the past half century, there have been outstanding advances in breast cancer management which have led to early detection of disease and the development of effective treatments, resulting in significant reduction in breast cancer mortality and

30 improved clinical outcomes for women living with the disease. Breast cancer treatment commonly involves surgical removal of the tumour, but this is ineffective if cancer cells have escaped from their primary site. A wide spectrum of clinical, pathologic, and molecular factors are routinely used to categorise patients in order to assess prognosis and determine the more effective therapeutic options. These include age, axillary node status, tumour size, histological grade, hormone receptor status, and HER2 status.

1.1.3 Molecular classifications of breast cancer

In recent years, several gene expression microarray studies have categorised breast cancer into four distinct molecular subtypes beyond the traditional hormone receptor- positive and hormone receptor-negative types: the luminal A, luminal B, human epidermal growth factor receptor-2 (HER2), and basal-like types (Schnitt, 2010,

Sørlie et al., 2001). These breast cancer molecular subtypes differ with regard to their gene expression profiles, clinical features, prognosis, and response to treatment, as summarised in Table1.1. Luminal-type breast cancers express the oestrogen receptor (ER) and associated genes and can be targeted with hormonal therapies. Subtype A is less aggressive and less sensitive to chemotherapy than subtype B. HER2 breast cancers are characterised by overexpression of HER2, with amplification of the HER2 gene. They are more aggressive than HER2 negative breast cancers, but can be targeted with the monoclonal antibody trastuzumab and are highly sensitive to chemotherapy. Basal-like breast cancers do not usually express ER, progesterone receptor (PR) and HER2, hence referred to as triple- negative. This type of breast cancer does not respond to hormonal therapies or

HER2-targeted therapies. They are highly proliferative and usually have a poor

31 prognosis, although they are highly sensitive to chemotherapy (Sørlie et al., 2001,

Schnitt, 2010, Allison, 2012, Zhang et al., 2014a).

Molecular subtype Luminal Her2 Basal High expression of hormone High expression of High expression of receptors and ER-related HER2 and Her- basal epithelial genes, genes (luminal A>luminal B) associated genes basal cytokeratins Gene expression Low expression of Low expression of ER pattern ER and and associated genes associated genes Low expression of HER2

~75% of invasive breast ER/PR negative Most ER/PR and cancers ER/PR positive HER2 negative (‘triple negative’) BRCA1 dysfunction A: Lower- B: Higher- Clinical features grade ER+ grade ER+ Low Ki67 High Ki67 Some overexpress HER2 Prognosis A: Good B:Intermediate Worse Worse Respond to endocrine Respond to No response to therapy (but response to trastuzumab endocrine therapy or tamoxifen and aromatase (Herceptin) trastuzumab inhibitors may be different (Herceptin) Treatment for luminal A and luminal B) Respond to response anthracycline- Appear to be sensitive Response to chemotherapy based to platinum-based variable (greater in luminal chemotherapy chemotherapy and B than in luminal A) PARP inhibitors

Table 1.1: Major molecular subtypes of breast cancer determined by gene expression profiling (Adapted from (Schnitt, 2010)).

These four different breast cancer subtypes represent biologically distinct diseases and are managed accordingly in the clinic. The therapeutic choices for breast cancer depend on the type of cancer, the menopausal status and stage of disease.

Endocrine therapies, such tamoxifen or aromatase inhibitors, remain the first-line treatment of hormone-sensitive metastatic breast cancer and as adjuvant therapy for early breast cancer in patients with hormone-receptor-positive tumours (Buzdar,

2003, Jaiyesimi et al., 1995); whereas, HER2 monoclonal antibody trastuzumab represents as the first-line treatment for patients with HER2-overexpressing

32 metastatic breast cancer (Vogel et al., 2002). The use of endocrine therapy or monoclonal antibody therapy has significantly reduced the number of deaths from breast cancer over the past decades. However, the majority of these patients will eventually develop resistance to endocrine or monoclonal antibody treatment (Ali and Coombes, 2002, Lu et al., 2001, Chang, 2012). In this case, and in patients with triple-negative, cytotoxic chemotherapy remains the mainstay of treatment in the metastatic setting. Altogether, it seems that most breast cancer patients of all subtypes receive chemotherapy as a part of the treatment process.

1.2 Chemotherapy

In the clinical treatment of breast cancer, chemotherapy is one of the most common treatments for the disease. Because cancer cells uncontrollably divide and are able to invade or spread to other tissues of the body, most chemotherapy drugs typically target rapidly dividing cells and function by interfering with the cell division process.

Chemotherapy drugs can be divided into several groups based on how they work to eliminate cancer cells and their chemical characteristics (Helleday et al., 2008).

Among the traditional cytotoxic chemotherapeutic drugs, anthracyclines and taxanes have emerged as the most powerful treatments in breast cancer (Kröger et al.,

1999).

1.2.1 Anthracyclines

Anthracyclines (also known as anthracycline antibiotics), including doxorubicin

(DOX) and epirubicin, are an important class of anti-tumour drugs with a broad spectrum of activities in human cancers, and only a few cancers (e.g. colon cancer) are unresponsive to them. DOX is one of the first anthracyclines that were originally isolated from the pigment-producing Streptomyces peucetius early in the 1960s. Its

33 chemical characteristics are composed of a four ring structure (rings A-D) and the aglycon chromophore, which is linked to an amino sugar (daunosamine) via a glycoside bond. DOX is one of the most widely used chemotherapeutic agents in the standard treatment of a wide range of solid tumours including breast, lung, gastric, thyroid and ovary carcinomas, as well as leukaemia (Minotti et al., 2004).

Unfortunately, its efficacy in treating cancer is limited by a cumulative dose- dependent cardiotoxicity, which causes congestive heart failure and several other side-effects to healthy tissue (Von Hoff et al., 1979, Ewer and Ewer, 2015).

Therefore, the modification of this drug to generate new analogues with improved activity and lower toxicity has been sought.

Over subsequent years with numerous attempts to develop and generate better anthracycline analogues, epirubicin, an epimer of DOX, have been approved for clinical use for treating a broad range of malignancies including breast, lung, gastric, and ovary carcinomas, and leukaemia (Minotti et al., 2004). The chemical structure of epirubicin differs from doxorubicin only at the spatial orientation of the C-4 hydroxyl group on the sugar (Figure 1.1). This slight change contributes to the faster elimination rate and reduced cardiotoxicity of epirubicin, making it preferential to

DOX.

Figure 1.1: Structures of the anthracyclines, doxorubicin and epirubicin. Red arrows indicate structural modifications in epirubicin compared with doxorubicin (axialto-equatorial epimerisation of the hydroxyl group at C-4′ in daunosamine. Despite the extensive and long-standing clinical use of anthracyclines, their precise mechanisms responsible for the anti-proliferative and cytotoxic effect are not completely understood. There are two main proposed mechanisms by which anthracyclines act in the cancer cell: (1) intercalation into DNA, leading to inhibition of DNA synthesis, and the initiation of DNA damage via the disruption of topoisomerase II-mediated DNA repair and (2) generation of free radicals, thus adversely altering cell membrane, proteins, and DNA (Minotti et al., 2004, Thorn et al., 2011, Gewirtz, 1999). These processes are able to render cancer cells to undergo apoptosis, mitotic catastrophe, or senescence.

DNA Topoisomerase II (TOP2) is likely to be one of the primary targets for the activity of the anthracycline antibiotics. DNA Topoisomerase II is an essential nuclear enzyme which regulates the topological state of the DNA without any alteration of deoxynucleotide structure and sequence. It can cause transient DNA double-strand breaks (DSBs) that are resealed after changing the twisting status of the double helix. This activity plays a critical role in modulating the supercoiling of the DNA double helix according to the cell cycle phase and transcriptional activity.

Anthracyclines act as topoisomerase II poisons. They stabilise a reaction in which

DNA strands are cut and covalently linked to tyrosine residues of topoisomerase II, eventually preventing DNA resealing and inducing the accumulation of permanent

DNA damage. The topoisomerase II-mediated DNA damage is then followed by growth arrest in G1 and G2 and subsequent the activation of programmed cell death

(Minotti et al., 2004, Gewirtz, 1999, Guano et al., 1999).

1.2.2 Paclitaxel

Paclitaxel, usually called by its brand name Taxol, is a commonly used microtubule- binding compound for treating advanced breast, ovarian, and lung cancers. It was originally isolated from the bark of Taxus brevifolia (pacific yew tree) in the 1960s

(Verweij et al., 1994, Wani et al., 1971). Microtubules are composed of heterodimers of α-tubulin and β-tubulin. Paclitaxel acts against cancer cells by binding to the β subunit of tubulin in microtubules, which stabilise and promote microtubule polymerisation, thereby preventing their depolymerisation and interrupting normal microtubule dynamics. Because microtubules are an essential part of the mitotic microtubule spindle apparatus during mitosis, this action results in cell cycle arrest in

G2/M phase and, ultimately, cell death through apoptosis, senescence or mitotic catastrophe process (Gornstein and Schwarz, 2014). Several studies have indicated that paclitaxel-arrested cells are in metaphase and certain concentrations of paclitaxel induce multipolar spindles (Weaver, 2014). Paclitaxel may also interfere with other cellular functions in which microtubules play a critical role, such as maintenance of cell shape, motility, and cellular transport (Goble and Bear, 2003).

Figure 1.2: Molecular structure of Paclitaxel. The chemical structure of paclitaxel is composed of four conjoined rings. The side chain of paclitaxel also holds functional groups that are essential for its anti-tumour activity.

1.3 Apoptosis and non-apoptotic deaths in treatment response

The cytotoxic treatment strategies of conventional chemotherapy have relied on the assumption that complete cellular destruction of tumours optimises the potential for patient survival (Ewald et al., 2010a, Kahlem et al., 2004). Genomic DNA damage is the pivotal cellular target of anti-cancer agents, which triggers programmed cellular response. Apoptosis has long been considered to be the principal mechanism of programmed cell death in response to chemotherapy. However, accumulating evidence suggests that in addition to apoptosis, necrosis, autophagy, mitotic catastrophe and senescence have also been identified as potential outcomes of cancer treatment (Figure 1.3). In particular, the fact that many cancer cells have inactivated their apoptotic programme cell death pathway during tumorigenesis, which could make non-apoptotic cell death pathway an alternative outcome for cancer therapy. The different resulting mechanisms of cytotoxic agents-induced cell death are likely determined by the mechanism of action of the drug, the dosing regimen used, the genotype of the cell within the tumour as well as the molecular status of cell-cycle checkpoints (Morse et al., 2005, Okada and Mak, 2004). The better examination of what we know about chemotherapy-induced cell death is thus crucially important in the light of new understanding about non-apoptotic cell death signalling pathways. If we can specifically activate molecules that play a critical role in mediating the complexity of cell death outcomes, we possibly can succeed in more effective and less toxic chemotherapeutic treatments (Ricci and Zong, 2006). In this study, I focus on two distinct crucial non-apoptotic mechanisms: senescence and mitotic catastrophe, which are usually triggered in cancer cells in response to chemotherapeutic drugs.

Figure 1.3: Chemotherapeutic agents activate various signalling pathways that can lead to different forms of cell death (Adapted from (Ricci and Zong, 2006)).

1.3.1 Cellular senescence

Cellular senescence, the irreversible growth arrest of cells in G1 phase of the cell cycle, is crucial for normal healing in damaged tissues, ageing and tumour suppression which prevents the growth of cells at risk for neoplastic transformation

(Rodier and Campisi, 2011). Cellular senescence was originally characterised by

Hayflick as a signal transduction process that limits cell proliferation of normal human cells in culture (Hayflick, 1965). Cells that undergo senescence cannot proliferate even if stimulated by mitogen, but they remain metabolically active and display morphological changes, such as enlarged and flatted cell shape, increased granularity and a vacuole-rich cytoplasm (Campisi, 2001). In addition to these classic morphological characteristics, there are a number of cellular biomarkers that typically associated with the senescent phenotype. The most widely used biomarker to identify senescent phenotype in both in vitro and in vivo is the increase in

“senescence-associated β galactosidase” (SA-βgal) activity, which involves a simple histochemical staining procedure at pH 6.0 to detect the expanded lysosomal compartment in the perinuclear region of aged cells (Dimri et al., 1995b). With advances in the biochemical characterisation, several other molecular senescence biomarkers have also been described, including high expression levels of p16INK4A

(p16) and p21CIP1/WAF1 (p21) (Dimri et al., 1995a, Serrano et al., 1997), senescence- associated heterochromatic foci (SAHF), which are nuclear DNA domains densely stained by DAPI, DNA segments with chromatin alterations reinforcing senescence, and the senescence-associated inflammatory cytokine secretion (Itahana et al.,

2013, Zhang and Yang, 2011b). This permanent growth arrested state in normal human somatic cells is referred to as ‘replicative senescence (RS)’. RS is mainly trigged by a combination of two main factors. The first factor is the progressive telomere shortening, which provokes a persistent DNA damage response (DDR)

(D'Adda Di Fagagna et al., 2003). Telomeres become gradually shorter after each cell division and are eventually too short to allow the cell to divide, resulting in cellular senescence or apoptosis (Harley et al., 1990, Fagagna et al., 2003). The

39 second factor is the activation of tumour-suppressive signals, including activation of the p53 and the p16 (also known as INK4A–RB pathways) (Collado et al., 2007).

In addition to replicative senescence, more recent studies which build on the phenomenon of replicative senescence in normal cells approaching the limit of their proliferative potential have identified a comparable senescence-like arrest as a component of the tumour cell response to chemotherapeutic drugs and radiation.

This response, which has been termed “premature senescence”, or “accelerated senescence”, can be induced by other types of stress such as oncogene activation

DNA damage, oxidative stress or chemotherapy treatment (Shay and Roninson,

2004, Kuilman et al., 2010).

Oncogene-induced senescence (OIS) has been identified in non-malignant human tissues as one mechanism of tumour suppression. The oncogenic protein Ras and its effectors, including activated mutant RAF, MEK, and BRAF, have been reported to trigger senescence (Braig and Schmitt, 2006, Zhang and Yang, 2011b). This form of senescence is similar to replicative senescence; however, OIS is much more rapid and independent of dysfunctional telomeres. One of the hallmarks shared by cells undergoing replicative senescence and OIS is the critical involvement of the p53 and p16INK4A–RB pathways (Ben-Porath and Weinberg, 2005). For example, benign melanocytic nevi (skin moles) result from the increased activity of the mutant oncogene v-raf murine sarcoma viral oncogene homologue B1 (BRAF). After increased proliferation and growth, melanocytes arrest, increase the expression of p16INK4A, and stain positive for senescence-associated acidic β-galactosidase (SA-β- gal) activity (Michaloglou et al., 2005, Ewald et al., 2010a). It is now becoming increasingly clear that oncogene induced senescence (OIS) is a critical endogenous

40 protective barrier against neoplastic transformation and that senescence in tumours may indicate a more benign or favorable outcome.

As mentioned above, tumour progression involves inhibiting crucial mediators of cellular senescence. However, it does not mean that transformed cancer cells have completely lost their capacity to undergo senescence. This response can still be activated by radiation or genotoxic chemotherapy. Current research suggests that therapy-induced senescence (TIS) represents a novel form of cellular senescence that may provide an effective approach to induce a persistent growth inhibitory response in both early- and late-stage cancers while limiting toxicity (Ewald et al.,

2010b, Schmitt et al., 2002, Xue et al., 2007). TIS can be induced in immortal and transformed cancer cells by selected anticancer compounds or radiation, and it shares many similarities to OIS (e.g. cellular biomarkers) (Zhang and Yang, 2011a,

Yang et al., 2012, Rebbaa et al., 2003). TIS has been shown to be a relevant factor in determining treatment outcome for breast and lung cancers (te Poele et al., 2002,

Roberson et al., 2005). Moreover, the effectiveness of senescence-inducing drugs could also be a potential alternative approach to treat tumours that are resistant to apoptosis-based therapies. Ewald et al. recently screened and identified diaziquone

(AZQ), a DNA alkylating agent, as a promising senescence-inducing compound in prostate cancer cells by using a semi-automated high-throughput method. They also revealed that Skp2 participates in regulating TIS in cancer cells (Ewald et al., 2009,

Ewald and Jarrard, 2012). Similarly, Chan et al. have also shown that a Skp2 inhibitor, which selectively inhibits Skp2 E3 ligase activity, exhibits potent anti-cancer activities in multiple animal models and cooperates with chemotherapeutic drugs to reduce cancer cell progression by triggering p53-independent cellular senescence

(Chan et al., 2013). In addition, a growing number of specific inhibitors that promote

41 senescence response, such as mTOR inhibitor (Everolimus) (Wall et al., 2013),

CDK4 inhibitor (PD0332991) (Puyol et al., 2010), Aurora A kinase inhibitor

(MLN8054) (Huck et al., 2010), MDM2 inhibitor (Nutlin3) (Arya et al., 2010, Polański et al., 2014, Efeyan et al., 2007), PTEN inhibitor (VO-OHpic) (Alimonti et al., 2010), and histone deacetylase (HDAC) inhibitors (Pazolli et al., 2012), have also been shown to have a relevant beneficial for cancer treatment .

However, it remains unclear what determines the choice of a cell to respond to the treatment by either triggering apoptosis or senescence (Childs et al., 2014). It could depend on exposure time and the dosages of the chemotherapeutic agents used.

The lower doses induce senescence, whereas the higher doses induce apoptosis

(Chang et al., 1999). For example, in prostate cancer cell lines, the lower 25 nM doses of doxorubicin induce senescence, whereas 250 nM of doxorubicin triggers apoptosis (Schwarze et al., 2005, Ewald et al., 2009). Moreover, accumulating evidence indicates that TIS reduces toxicity-related side effects, increased tumour- specific immune activity and improved treatment outcome (Schmitt et al., 2002, Xue et al., 2007). Recently, the ability of cancer cells to overcoming TIS has been proposed as one mechanism behind cancer recurrence and drug resistance.

1.3.2 Mitotic Catastrophe

The other anti-proliferative response of tumour cells is mitotic catastrophe, a mode of cell death characterised by the occurrence of aberrant mitosis or mis-segregation of chromosomes leading to the formation of giant interphase cells with abnormal nuclear morphology and multiple micronuclei, which are morphologically distinct from

42 apoptosis, necrosis and autophagy (Table 1.2) (Vitale et al., 2011, Weaver and

Cleveland, 2005). Centrosomes have an important role in the formation of bipolar mitotic spindles, which are essential for accurate chromosome segregation. Mitotic catastrophe also takes place as a result of centrosome over duplication, which leads to consequent entry into mitosis with multiple spindle poles (Fragkos and Beard,

2011, Sato et al., 2000) or failure of centrosomes to undergo duplication with the consequent failure of chromosome to segregate (Cogswell et al., 2000). In mammalian cells, particularly in cancer cells, mitotic catastrophe is mainly related with defective cell cycle checkpoints, including both the DNA replication checkpoint and the spindle assembly checkpoint, and cellular damage (Castedo et al., 2004).

The G2 checkpoint of the cell cycle is responsible for blocking cell division when a cell has sustained an insult to DNA. DNA damage activates various cellular responses, including cell cycle arrest, DNA repair, or cell death. If the G2 checkpoint is defective, a cell can enter mitosis inappropriately, before DNA replication is complete or DNA damage has been repaired. This aberrant mitosis renders the cell to undergo death by mitotic catastrophe. In addition, mitotic catastrophe can be triggered by agents influencing the stability of microtubules, various classes of chemotherapeutic drugs and ionising radiation, but the pathways of abnormal mitosis differ depending on the nature of the inducer and the status of cell-cycle checkpoints

(de Bruin and Medema, 2008, Roninson et al., 2001). For example, paclitaxel causes an abnormal metaphase in which the sister chromatids fail to segregate properly

(Jordan et al., 1996). CDK1 activation is prolonged abnormally in paclitaxel-treated cells, which undergo cell death by mitotic catastrophe. Mitotic catastrophe is driven by a number of molecular players, in particular, cell cycle specific kinases (such as

CDK1, PLK1 and Aurora kinases), cell cycle checkpoint proteins, including survivin,

43 p53, caspases and members of the Bcl-2 family (Castedo et al., 2004). Elucidation of the genes and regulatory mechanisms that determine different aspects of treatment- induced mitotic catastrophe may help in improving the efficacy of anti-cancer therapy, providing opportunities for the development of new drugs.

Type of cell death: Apoptosis Senescence Mitotic catastrophe Blebing, membrane Flattening, increase Cell integrity maintained in granularity and morphology cell size Nuclear fragmentation, Accumulation of Micronuclei, mis- Morphological chromatin heterochromatin segregation of characteristics Nucleus condensation, DNA (SAHF) foci, chromosomes, nuclear laddering prolonged γ-H2AX fragmentation foci Fragmentation Cytoplasm (formation of apoptotic bodies) SA-βgal activity Abnormal CDK1/Cyclin Biochemical features B activation, Caspase dependent Caspase independent Caspase-activity SA-βgal staining, Visualisation of assays, TUNEL assay, Clonogenic assay, multinucleated cells, Annexin V staining, Flattened cell Detection of mitotic detection of changes in morphology, marker MPM2 Detection mitochondrial detection of membrane potential increased p53, p21, p16 INK4A and p19ARF, detection of elevated SASP factors (including IL-6 and IL-8) Table 1.2: Cell death pathway characteristic (Adapted from (Okada and Mak, 2004)).

1.4 General mechanisms of drug resistance in breast cancer

Despite advances in detection, adjuvant therapy, and the understanding of the molecular biology of breast cancer, about 30% of patients with early-stage breast cancer develop recurrent and metastatic disease (Pisani et al., 2002, Moreno-Aspitia

44 and Perez, 2009). Resistance to chemotherapy represents the main obstacle against the successful management of breast cancer. It is thought to cause treatment failure in over 90% of patients with metastatic cancer. Drug resistance can be divided into two main categories: pre-existent (intrinsic resistance), or induced by drugs

(acquired resistance). Acquired resistance is a particular problem, as cancers not only become resistant to the drugs originally used to treat them, but may also become cross-resistant to other anticancer agents with different structure and mechanism of action (Longley and Johnston, 2005). Drug resistance can be mediated by a number of different biochemical mechanisms (Figure 1.4). These include: (1) elevated expression of drug efflux pumps such as permeability glycoprotein (P-gp) (Ambudkar et al., 1999); (2) reduced drug influx; (3) increased drug inactivation or detoxification; (4) alteration of drug targets; (5) enhancement of

DNA repair capacity; (6) reduced ability to undergo apoptosis or non-apoptotic cell death; or (7) mutations of cell cycle checkpoints leading to overexpression of anti- apoptotic genes (Broxterman et al., 2003, Holohan et al., 2013, Gottesman, 2002).

Although chemotherapy resistance is a main problem, an understanding of these biochemical changes can provide potential targets to overcoming drug resistance in human cancer.

Chemotherapy resistance also seems to arise from the expression of proteins underlying a specific mechanism of action of each anti-cancer drug. For example, paclitaxel functions by binding to β-tubulin. Therefore, paclitaxel resistance has been linked to a wide variety of mechanisms including molecular mutations in the target molecule (β-tubulin) leading to alterations in microtubule dynamics (Minotti et al.,

1991, Giannakakou et al., 1997, Ganguly et al., 2011), alteration of the expression pattern of microtubule associated proteins (MAPs) (Rouzier et al., 2005), defective

45 spindle assembly checkpoint (SAC) (Anand et al., 2003), changes in cell death regulatory proteins (Ferlini et al., 2003, Wang et al., 2005b), and overexpression of multidrug resistance (MDR-1) gene (Wind and Holen, 2011, Murray et al., 2012).

Resistance to anthracyclines develops via a wide variety of mechanisms including mutations in topoisomerase II (the molecular target of the anthracyclines) (Bugg et al., 1991, Ganapathi and Ganapathi, 2013), overexpression of DNA repair proteins

(Monteiro et al., 2013, Kwok et al., 2010a, Millour et al., 2011, Park et al., 2012), alterations in cell death signalling (Fan et al., 1994), and increased drug efflux mediated by overexpression of ATP-binding cassette transport proteins, such as P- glycoprotein (Pgp), multidrug resistance (MDR) protein (MRP-1), and breast cancer resistance protein (BCRP) (Chien and Moasser, 2008, Minotti et al., 2004).

Although multiple mechanisms have been identified to be responsible for the resistance to a wide range of chemotherapeutic agents, a deeper understanding of how the exact mechanism involved in the molecular pathways underlying drug sensitivity and resistance is still required. These will eventually lead to the development of novel therapeutic strategies which will have the potential to improve the efficacy of current treatment of breast cancer with drug resistance characteristics.

Figure 1.4: Molecular mechanisms of anti-cancer drug resistance. Different mechanisms drive cancer cells to become resistant to one or more cytotoxic anticancer drugs known as multidrug resistance (MDR) significantly suppress the effectiveness of cancer chemotherapy. Potential factors for MDR include: increased drug inactivation or detoxification; alteration of drug targets; enhancement of DNA repair capacity; reduced ability to undergo apoptosis or non-apoptotic cell death; decreased drug uptake; mutations of cell cycle checkpoints; overexpression of efflux pumps such as P- glycoprotein (P-gp) multidrug resistance protein family members (MRPs), and breast cancer resistance protein (BCRP) (Adapted from (Kapse-Mistry et al., 2014)). The red crosses represent inhibitory processes.

1.5 DNA Damage Response (DDR)

Genomic DNA is the most critical component of cells. The integrity of DNA in each cell is continually damaged through a combination of endogenous processes (e.g. stalled forked during replication, telomere erosion, the generation of reactive oxygen species during oxidative metabolism) and exogenous factors (e.g. radiation and genotoxic compounds) that alter the sequence or chemical composition of the DNA

(Jalal et al., 2011). These lesions may result in the formation of single-strand or double-strand DNA breaks, bulky adducts, intrastrand and interstrand crosslinks and breakdown of the replication forks. These are dangerous events because they compromise the structural stability of chromosomes and induce genomic instability, which is one of the hallmarks of cancer. In order to counteract DNA lesions and maintain genomic stability, cells can stimulate and amplify various biochemical pathways, collectively termed as the DNA damage response (DDR) (Jackson and

Bartek, 2009). In general, the cellular DNA damage response occurs through an integrated sensing and signalling network, which is composed of a number of gene products, including sensors, transducers and effectors. DSBs are initially detected by

DNA damage sensor molecules, which trigger the activation of downstream transducing kinases. These transducers amplify damage signals by phosphorylation of effector proteins, which in turn regulate cell cycle progression, DNA repair and apoptosis (Jackson and Bartek, 2009). Immediately after DSBs, the DSB sites are recognised by the MRE11-RAD50-NBS1 (MRN) complex which recruits ataxia telangiectasia-mutated (ATM), a key element in the DNA damage response pathway, to the damage site. This event also activates ATM by inducing its Ser139 autophosphorylation (Bakkenist and Kastan, 2003, Lee and Paull, 2005). Then, activated

ATM phosphorylates histone H2AX on Ser139 (Burma et al., 2001). This

48 phosphorylated form, known as γH2AX, induces the recruitment of additional DNA damage responsive proteins including BRCA1, MRN complex and 53BP1 to the

DNA damage sites and results in the formation of DNA damage foci. Activated ATM also functions by phosphorylating and activating a number of key proteins, involved in DNA damage repair, DNA replication, cell cycle checkpoint arrest and apoptosis, such as p53, NBS1, BRCA1, CHK1, CHK2, E2F1, MDM2, and SMC1, to ensure the damaged cells do not continue dividing until the DNA damage is repaired

(Darzynkiewicz et al., 2009, Kurz and Lees-Miller, 2004, Sulli et al., 2012)

(Figure1.5). Once a DNA lesion is repaired, DDR foci are disassembled. This is probably due to the action of both chromatin remodelling machines and the dephosphorylation of γH2AX by dedicated phosphatases. Therefore, promptly repaired lesions are expected to display transient and relatively small foci, whereas severe or irreparable DNA breaks will induce more protracted DDR signalling and increased γH2AX foci spreading and consequently produce visibly bigger foci. DSBs that cannot be repaired (for example, uncapped telomeres) cause constitutive DDR signalling, prolonged p53-dependent growth arrest and eventually an irreversible senescence (d'Adda di Fagagna, 2008).

Figure 1.5: The model of the ATM signalling pathway in response to induction of DSBs. When DSBs induced by radiation or chemotherapy are present, ATM is recruited and become activated via an interaction with the main DNA damage sensor MRN complex, Activated ATM then phosphorylates H2AX (also known as γH2AX). In addition, ATM also phosphorylates BRCA1, 53BP1, and MDC1 as well as checkpoint protein Chk2. This process is aimed to stop cell cycle progression and to activate p53, a downstream effector responsible for DNA repair. DDR-mediated cellular outcomes could be transient cell cycle arrest followed by repair of DNA damage and resumption of cell proliferation; cell death by apoptosis; or cellular senescence caused by the persistence of unrepaired DNA DSBs (Adapted from (Sulli et al., 2012)).

1.5.1 DNA double strand break repair

Cells possess a number of distinct, but partially compensatory, DNA repair mechanisms, each addressing a specific form of DNA breaks (Jackson and Bartek,

2009). DNA double-strand breaks (DSBs) are considered the most deleterious form of DNA lesions, which can lead to genomic instability, cancer, cell death, or cellular senescence (Vilenchik and Knudson, 2003). Therefore, the activation of cellular checkpoints and effective repair in response to DSBs are the critical barriers to prevent malignant transformation. DNA DSBs can be caused by many different environmental factors, including reactive oxygen species, ionising radiation and certain chemotherapeutic agents, such as anthracyclines and topoisomerase inhibitors. Alternatively, DSBs can result from endogenous factors, especially during

DNA replication process (Khanna and Jackson, 2001). Moreover, DSBs naturally occur at chromosome ends, related to human cell ageing and replicative senescence

(D'Adda Di Fagagna et al., 2003). Persistent or incorrectly repaired DSBs can cause chromosome translocations and genomic instability, potentially leading to multiple cancers and immunodeficiency. The prevalence of DNA DSBs in cancer due to uncontrolled cell proliferation and defects in repair opens up a potential therapeutic window for cancer treatment. Therefore, targeting the proteins that is essential for the repair of DSBs with chemotherapeutic agents has become a potential strategy for cancer therapies.

In mammalian cells, there are two main pathways responsible for DSBs, namely homologous recombination (HR) pathway and non-homologous end joining (NHEJ) pathway (Figure 1.6). The choice of cells to trigger either NHEJ or HR pathway is dependent on the phase of the cell cycle. The HR pathway is active specifically in S and G2 cell cycle phases, the stages when a sister chromatid is available as a template for targeted HR (Johnson and Jasin, 2000). This process allows for error- free repair of DNA damage. By contrast, the NHEJ pathway is an error-prone DNA repair pathway, which processes and ligates the two broken DNA ends directly and can be activated in all cell cycle phases (Choudhury et al., 2009). NHEJ is often mutagenic because deletions or insertions can be induced at the repair sites.

In NHEJ, both ends of the DNA break are first bound by the Ku70/Ku80 heterodimer, which recruits and activates the DNA-dependent protein kinase catalytic subunit

(DNA-PKcs) to form the DNA-PK holoenzyme (Giffin et al., 1996). Broken DNA ends are then processed by the Artemis nuclease before being ligated by a multimeric complex consisting of X-ray repair cross-complementing protein 4 (XRCC4),

XRCC4-like factor (XLF) and DNA ligase IV. HR takes over if NHEJ is unsuccessful in re-joining the broken DNA ends or when the DSB is first recognised by the

MRE11-RAD50-NBS1 (MRN) complex rather than by Ku70/Ku80 (Symington and

Gautier, 2011).

Another important regulatory step that determines the choice between HR and NHEJ

DSB repair pathway is the process of DSB resection, which is required for HR but not NHEJ. Immediately after DSBs, the DSB sites are firstly recognised by the

MRE11-RAD50-NBS1 (MRN) complex. In conjunction with CtBP-interacting protein

(CtIP), the RECQ family helicases, and the nucleases EXO1 and DNA2, the MRN complex resects DSBs to generate short 3’-single stranded DNA (ssDNA)

52 overhangs, which are immediately coated with replication protein A (RPA) before being replaced by RAD51 with the help of the BRCA2-PALB2 complex that plays key roles in detecting and repairing interstrand cross-links, particularly at sites of stalled

DNA replication (Sartori et al., 2007, Zou and Elledge, 2003, Bernstein and

Rothstein, 2009). The RAD51 nucleofilament, together with various other HR factors, then searches and invades into the homologous template forming a displacement loop to initiate repair. The second of the broken chromosome is captured and anneals the complementary strand of the DNA molecules, leading to the formation of two Holiday junctions (HJs). After the process of DNA synthesis by DNA polymerases and DNA end ligation of both strands by Ligase I, the double HJ is then cleaved by DNA helicase and resolvase enzymes in order to complete repair

(Moynahan and Jasin, 2010, Krejci et al., 2012).

Figure 1.6: DNA double strand break repair mechanisms. In NHEJ, DNA DSBs are recognised by the Ku70/80 heterodimer which in turn recruits DNA-PK. The blunt ends of DNA breaks are processed by Artemis nuclease. The DNA-PK complex then phosphorylates and stimulated the NHEJ effector complex (Ligase IV/XRC44/XLF) that ligates the broken DNA. In HR, ATM is recruited to DSBs via an interaction with the MRN (Mre11/Rad50/Nbs1) complex. Once ATM becomes activated, it phosphorylates several DNA damage effectors. DSBs are resected forming ssDNA strands by various nucleases activities, such as MRE11, EXO1, DNA2 and CtIP. These ssDNAs are immediately coated with replication protein A (RPA) before being replaced by RAD51. The Rad51 nucleoprotein filaments then induce invasion into the undamaged sister strands forming a Holliday junction (HJ). HR is completed by new DNA synthesis and the double HJ is finally cleaved by HJ resolvase enzymes.

1.5.2 NBS1

NBS1 (NBN/Nibrin), the gene product of mutation in Nijmegen breakage syndrome

(NBS) patients, is the p95 component of the MRE11-RAD50-NBS1 (MRN) complex.

The MRN complex is one of the main players in cellular response to DSBs, involving damage sensing and activation of signaling pathways that control cell-cycle checkpoints in response to damage and repair (Stavridi and Halazonetis, 2005). The importance of NBS1 in DNA repair is indicated by systemic defects in Nijmegen breakage syndrome (NBS) patients including premature ageing, immunodeficiency and a high frequency of malignancies (Kobayashi et al., 2004). Moreover, NBS defective cells are hypersensitive to radiation, chromosomal fragility, and abnormal cell cycle checkpoint regulation, as well as having a high frequency of malignant phenotypes similar to those of ataxia telangiectasia mutated (ATM) (Saito et al.,

2013). The gene responsible for Nijmegen breakage syndrome, NBS1, was originally identified in 1998 (Carney et al., 1998, Varon et al., 1998). As shown in Figure 1.7, it contains 16 exons encompassing 50kb on chromosome 8q21 (Saar et al., 1997,

Varon et al., 1998, Kobayashi et al., 2004). Human NBS1 encodes a 754 amino acid-NBS1 protein, which is composed of three functional regions: the N-terminus; central region; and the C-terminus. The N-terminus of NBS1 contains a forkhead- associated (FHA) domain and two breast cancer C-terminus (BRCT) domains. The

FHA/BRCT domain directly interacts with histone γH2AX, the phosphorylated form of

H2AX in response to the presence of DSBs, then recruits MRE11 and RAD50, forming the MRN complex to the vicinity of the DSB sites (Figure 1.7). Moreover, it has been shown to interact with MDC1, γH2AX, TOPBP1, and WRN. The C-terminal motifs are indispensable for binding to MRE11, ATM, RAD18, and RNF20 (Saito et al., 2013). For the functional relationship between ATM and NBS1, NBS1 appears to

55 function as a downstream mediator of ATM in response to DNA damage. ATM specifically interacts with and phosphorylates NBS1 on specific residues, serine 278 and serine 343, in response to DNA damage (Figure 1.7). This process is essential for S phase checkpoint activation, formation of DNA damage foci and rescue of hypersensitivity to ionizing radiation (Zhao et al., 2000, Gatei et al., 2000, Lim et al.,

2000). However, recent studies revealed that the mutation of members of the MRN complex or NBS1 leads to impaired ATM activation. Furthermore, the binding of

NBS1 is critical for ATM activation to be fully functional in response to DNA damage; therefore, NBS1 acts both upstream and downstream of ATM (Cariveau et al., 2007,

Kang et al., 2005).

Figure 1.7: The structure and protein interaction sites of human NBS1/Nibrin. The NBS1 protein is composed of three main functional domains: the FHA/BRCT domain at the N-terminus; ATM- phosphorylated serine residues, S278 and S343, at a central region; and MRE11-binding region at the C-terminus. The N-terminal FHA/BRCT domain directly interacts with γ-H2AX, TopBP1 and MDC1, whereas the C-terminal motifs are indispensable for binding MRE11, ATM, RAD18, and RNF20 (Adapted from (Kobayashi et al., 2004)).

1.6 The cell cycle and cancer

The cell cycle, or cell division cycle, is defined as the tightly controlled series of events that take place in a cell leading to its division and duplication. Cell division consists of two consecutive steps, mainly characterised by DNA replication and segregation of replicated chromosomes into two separate cells. The typical mammalian cell cycle is divided into four distinct phases: gap1 (G1), gap2 (G2), DNA synthesis (S), and Mitosis (M). The G1 phase prepares the cell for DNA replication, and is followed by S-phase during which chromosomes are actively replicated.

Subsequently, the cell enters a further gap period (G2) prior to chromosome segregation and cytokinesis in M phase (Swanton and Jones, 2001). The G1, S and

G2 represent the interphase of a proliferating cell. Cells in G1 can, before commitment to DNA replication, enter a resting state called G0. Cells in G0 account for the major part of the non-proliferating cells in the human body and they can re- enter into G1 if appropriate signals are received (Vermeulen et al., 2003).

The M phase is itself composed of two main processes: mitosis, in which the sister chromatids are aligned along the equator of the cell and then splits into two identical daughter cells; and cytokinesis, in which the cytoplasm and its components are equally divided into those cells. Mitosis can be subdivided into four distinct phases: prophase, metaphase, anaphase and telophase (Figure 1.8).

7.5 µm Figure 1.8: The four main stages of mitosis in MCF-7 cells. Cells were immunostained with antibody against α-tubulin (Green), Cells were immunostained with antibody against α-tubulin (Green), γ-tubulin (Red) and Nuclei were stained with DAPI (Blue). Cells were visualised with Leica TCS SP5 (63X magnification).

Prophase The chromatin condenses into chromosomes by dehydrating and coiling and CDK1- CyclinB translocates to the nucleus.

Metaphase Chromosomes are attached to spindle microtubules and moved to align at the centre of the spindle in a “bi-oriented” configuration. Unattached kinetochores generate a checkpoint signal that prevents the metaphase/anaphase transition until all of the kinetochores are correctly bi-oriented.

Anaphase Once the mitotic checkpoint has been approved, the APC ubiquitin ligase is activated. Ubiquitin-dependent proteolysis of the inhibitor securin leads to activation of separase, a protease that cleaves the cohesin proteins that hold sister chromatids together. This allows the chromosomes to separate.

Telophase After the chromosomes reach the poles, CDK1 activity is inhibited by APC-mediated destruction of Cyclin B. The reduction in CDK1 activity allows for reformation of the new nuclear envelope, decondensation of chromosomes, dividing the cytoplasmic components and entry into G1.

1.6.1 Cell cycle checkpoints

The orderly progression of one cell cycle phase to another is crucial for ensuring faithful DNA replication and chromosome segregation, thereby preserving genetic stability of daughter cells. A complex network of cell cycle control mechanisms, called cell cycle checkpoints, are responsible for ensuring that DNA replication occurs at G1 point. The next checkpoint (G2) ensures that mitosis only starts after chromatids have been precisely replicated and DNA repair completed (Swanton and

Jones, 2001). The checkpoint during mitosis ensures that chromosome segregation is corrected. Cells can be temporarily arrested at cell cycle checkpoints to allow for

DNA damage to be repaired. Persistent checkpoint signalling may also result in activation of pathways leading to programmed cell death or senescence if damaged

DNA cannot be repaired (Pietenpol and Stewart, 2002). Defective cell cycle checkpoint regulation frequently results in genomic instability, gene mutations, chromosome damage, and aneuploidy/polyploidy, all of which can contribute to uncontrolled proliferation essential for cancer development (Pietenpol and Stewart,

2002). The transition from one phase of the cell cycle to the next is regulated by different cellular proteins. Key regulatory factors of the cell cycle are CDK/Cyclin complexes.

G1/S checkpoint (restriction point) is the first checkpoint located at the mid G1 phase, prior to entering S phase, making the critical decision of whether the cell should divide, arrest/delay division or enter a resting stage called G0. This checkpoint is regulated by the Rb protein (pRb), the product of the retinoblastoma tumour suppressor gene. To pass this restriction point, CDK4/6 and CDK2 form active complexes with CyclinD and Cyclin E, respectively. The activity of active

CDK4/Cyclin D and CDK2/Cyclin E complexes induce the phosphorylation of pRb,

59 leading to the inactivation of its function as a transcriptional repressor. When pRb is bound to E2F, the complex acts as a growth suppressor, prevents cell cycle progression and the cell remains activity in the G1 phase; however, once hyper- phosphorylated, pRb releases E2F. The release of E2F activates a number of S- phase genes, including Cyclin D, Cyclin E and Cyclin A, prior entering into the S phase (Pietenpol and Stewart, 2002).

CDK activity is negatively regulated by CDK inhibitors, which specifically inactivate

CDK/Cyclin complexes and thereby cause cell cycle arrest. These CDK inhibitors

(CKIs) are divided into two families: the INK4 and the Cip/Kip inhibitors. INK4 CKIs include p15INK4B, p16INK4A, p18INK4C, and p19INK4D. The second family, the CIP/KIP

CKIs, consist of p21CIP1/WAF1, p27KIP1 and p57KIP2. The INK4 family specifically inhibits

CDK4 and CDK6 activity by displacing the cyclins from CDKs during the early G1 phase of the cell cycle, while the CIP/KIP family can inhibit CDK activity during all phases of the cell cycle. Both families of CDK inhibitors play an important role in anti- proliferative signals by arresting cells in G1 to enable such diverse processes as repair of DNA damage, terminate differentiation and cellular senescence (Sherr and

Roberts, 1995). Accumulating evidence suggests that the signal for the G1/S checkpoint is unreplicated DNA rather than DNA damage. After passing the restriction point, pRb is maintained in a hyper-phosphorylated state by the sequential activities of CDK2/Cyclin A and CDK1/CyclinA, thereby ensuring cell cycle progression into G2 phase. During S and G2 phases, cells accumulate CDK1/Cyclin

B1 complexes in their inactive form due to the inhibitory phosphorylation by Wee1

(Russell and Nurse, 1987) and Mik1 kinase (Lundgren et al., 1991). CDC25C phosphatase is the main factor that activates CDK1/Cyclin B1 complex, leading to the onset of mitosis. Therefore, the CDK1/CyclinB1 complex is also known as M-

Phase-promoting factor (MPF) (Rhind and Russell, 2012, Pietenpol and Stewart,

2002). However, when cells encounter DNA damage in G2, the G2/M checkpoint stops the cell cycle in order to prevent the cell from entering mitosis. In response to

DNA damage during G2 phase, members of PI-3K family, including ATM, ATR and

DNA-PK become activated and start signal transduction pathways that regulate cell cycle progression and DNA repair (Canman and Lim, 1998). The ATM-dependent signalling induced by DNA damage triggers the activation of CHK1 and CHK2 kinases (Sanchez et al., 1997, Furnari et al., 1997, Matsuoka et al., 1998). The activated CHK1 and CHK2 then phosphorylate CDC25C. This phosphorylation promotes the binding of CDC25C with 14-3-3 adaptor proteins and inhibits the ability of CDC25C to activate CDK1/Cyclin B1, leading to cell cycle arrest at the G2/M transition. Several studies also revealed that CHK1 and CHK2 mediated phosphorylation of p53 play an important role in the stabilising DNA repair proteins in response to DNA damage (Shieh et al., 2000). Moreover, p53-dependent transcription elevates the CDK inhibitor p21CIP1, which interacts with CDK/Cyclin complexes to inhibit the phosphorylation of pRB. Hypo-phosphorylated pRB tightly binds to E2F, preventing E2F from mediating the biosynthesis of Cyclin B1 and

CDK1 (Pietenpol and Stewart, 2002). Both tumour suppressors, p53 and p21CIP1, have been shown to be necessary for maintaining the G2/M arrest in response to

DNA damage and to maintain genomic stability (Di Leonardo et al., 1994). Once

DNA damage is completely repaired, the DNA damage checkpoint is silenced so that cell cycle progression is allowed to resume by the upregulation and activation of

CDK1/CyclinB complex, leading to the onset of mitosis.

In mitosis, the spindle assembly checkpoint (SAC), also known as mitotic checkpoint, acts to maintain genomic stability by delaying the segregation of chromosome in

61 anaphase until sister chromatids are properly attached and bi-oriented on the microtubule spindle via their specialised protein structures, called kinetochores

(Lara-Gonzalez et al., 2012, Kops et al., 2005, Musacchio and Salmon, 2007). In the presence of unattached kinetochores the SAC is ‘on’ and anaphase is halted.

During early mitosis, unattached kinetochores contribute to the formation of the spindle assembly checkpoint (SAC), consisted of Mad2, BubR1, Bub3 and CDC20, leading to the CDC20-dependent inhibition of the anaphase-promoting complex/cyclosome (APC/C). Once all sister chromatids are aligned with their kinetochores attached to the spindle (metaphase), the SAC signal is silenced and releases CDC20. The free CDC20 then activates APC/C activity, resulting in the ubiquitination and degradation of Cyclin B and securin. The degradation of securin activates separase, which in turn cleaves a subunit of the cohesin ring structure,

Scc1. This event triggers the separation of sister chromatids and marks the initiation of anaphase. Meanwhile, the loss of Cyclin B also inactivates the master mitotic kinase CDK1/Cyclin B, initiating cytokinesis and the mitotic-exit programme

(Musacchio and Salmon, 2007, Lara-Gonzalez et al., 2012).

1.7 Kinesins and Cancer

The mitotic spindle is an effective and validated target for cancer chemotherapeutic drugs. Well-known anti-mitotic drugs, such as taxanes and vinca alkaloids, which target tubulin and interfere with mitotic-spindle microtubule dynamics, are clinically successful anticancer drugs for metastatic breast cancer, as well as lung and ovarian carcinoma. Unfortunately, these spindle poisons have limitations, such as the development of drug resistance and dose-limiting toxicities, making it necessary to find alternative ways for targeting the mitotic spindle. An important family of proteins that has been shown to be essential for mitosis and that is emerging as a potential molecular target for novel chemotherapeutic intervention is the kinesin superfamily

(Rath and Kozielski, 2012).

The kinesin (also known as KIF) superfamily is a group of molecular motor proteins that consume energy from ATP hydrolysis to move along microtubules (MT) toward plus end (Vale et al., 1985, Vale et al., 1996, Vicente and Wordeman). It is composed of two main functional domains: an ATP-hydrolysing motor domain and a tail domain. The motor domain is highly conserved among the different kinesin families and enables motor binding and stepping along microtubules by converting the chemical energy of ATP hydrolysis into a mechanical force (Goldstein and Philp,

1999). Because of their specialised structure, KIF proteins have been functionally linked to several essential cellular activities, including mitotic spindle formation and chromosome partitioning, and migration, as well as intracellular movements of organelles and vesicles (Rath and Kozielski, 2012, Hirokawa et al., 1998). In mitosis, the activities of KIFs on the spindle microtubules are precisely regulated to ensure that mitotic events are orchestrated in the correct order throughout mitosis.

Currently, at least 12 kinesins have been implicated in coordinating mitosis and

63 cytokinesis. KIFC1 (a member of the kinesin-14 family), Eg5 (a member of the kinesin-5 family), and KIF2A (a member of the kinesin-13 family) play roles in mitotic spindle dynamics; KIF2C (a member of the kinesin-13 family), KIF18 (a member of the kinesin-8 family), CENP-E (a member of the kinesin-7 family), KIF14 (a member of the kinesin-3 family), and KID (a member of the kinesin-10 family) are essential for chromosome alignment; KIF4A and KIF4B (members of the kinesin-4 family) is associated with anaphase spindle dynamics; and MKLP1 and MKLP2/KIF20A

(members of the kinesin-6 family) are required for cytokinesis (Zhu et al., 2005, Rath and Kozielski, 2012). Besides the role in mitosis, the impacts of kinesins in tumour development and progression, as well as in the development of drug resistance, are beginning to emerge in more detail (Rath and Kozielski, 2012).

1.7.1 KIF20A

Kinesin family member 20A (KIF20A), also known as RAB6KIFL/mitotic kinesin-like protein 2 (MKLP2), is a newly identified member of the Kinesin superfamily proteins

(KIFs), which was originally identified to contribute to Golgi apparatus dynamics via interaction with the GTP-bound form of Rab6 during interphase (Echard et al., 1998).

KIF20A has previously been reported to accumulate in mitotic cells, where it localises to the midzone of the spindle during anaphase and at the midbody during telophase (Hill et al., 2000, Fontijn et al., 2001). KIF20A is believed to be essential for cell cycle regulation during successful cytokinesis, and its depletion causes a failure of cleavage furrow ingression and cytokinesis (Hill et al., 2000, Neef et al.,

2003, Fontijn et al., 2001). KIF20A is required to relocate and regulate chromosomal passenger protein Aurora B and the mitotic regulator PLK1 to the central spindle during anaphase (Gruneberg et al., 2004, Neef et al., 2006). Moreover, a study in

Drosophila also showed that Subito, which is an orthologue of KIF20A in mammalian

64 cells, localises to the spindle midzone during anaphase and is also required for localisation of PLK1, INCENP and Aurora B, which are essential for cytokinesis

(Cesario et al., 2006). Microinjection of antibodies against KIF20A led to multinuclear cells, further supporting a critical role of KIF20A in cytokinesis (Taniuchi, Nakagawa et al. 2005). KIF20A expression appears to be tissue specific. It is widely expressed in human fetal tissues. It is also expressed in the adult during haematopoiesis and in various proliferating tissues, and is abundantly expressed in adult thymus, bone marrow and testis (Lai et al., 2000). By contrast, there is no expression of KIF20A in adult quiescent human liver cells, suggesting that KIF20A contributes to both normal and pathological proliferation as well as cancer progressiveness in human cells

(Gasnereau et al., 2012). Recent studies revealed that KIF20A is overexpressed in pancreatic ductal adenocarcinoma (PDAC) cells and its downregulation inhibits cell growth in pancreatic and gastric cancer (Yan et al., 2012). The role of KIF20A in mitotic spindle formation and cytokinesis has been well studied in mitosis. However, little is known about the transcriptional regulation of KIF20A and it remains to be elucidated whether KIF20A plays a role in carcinogenesis or anti-cancer drug resistance.

1.8 FOXM1

1.8.1 FOXM1 gene and its protein structure

Transcription factors are proteins that bind specifically to defined DNA sequences to promote gene expression. FOXM1, also known as Trident (in mouse), WIN or INS-1

(in rat), MPP2 (partial human DNA), and HFH-11 (in human), belongs to the superfamily of forkhead transcription factors. The forkhead family comprises a large number of transcription factors which share an evolutionally conserved ‘winged helix’

DNA binding domain (Korver et al., 1997a). FOXM1 regulates the expression of its target gene via its binding to sequence-specific motifs on DNA sequence

TAAA(C/T)A (Korver et al., 1997b) and activates cell proliferation and regulatory cell cycle-associated genes (Laoukili et al., 2007). Recent research has revealed that

FOXM1 is a direct transcriptional target of FOXO3a, an important downstream effector of PI3K-AKT-FOXO signalling pathway. Activated FOXO3A antagonizes

FOXM1 activity by competitively binding to the same target genes, which are involved in cancer initiation, progression and chemotherapeutic sensitivity and resistance (Lam et al., 2013, Karadedou et al., 2012).

As shown in Figure 1.9A, the human FOXM1 gene consists of 10 exons, spanning approximately 25 kb on the 12p13.3 chromosomal band. Two exons, named exon

Va and Vlla, can be alternatively spliced, to provide three distinct isoforms, including

FOXM1a, FOXM1b, and FOXM1c (Lam et al., 2013, Koo et al., 2012). The FOXM1a isoform harbours both Va and Vlla exons and is transcriptionally inactive due to the disruption of its transactivation domain by the inhibitory exon Vlla. However,

FOXM1a can still interact with DNA. Therefore, it is considered to function as a dominant negative regulator of other FOXM1 isoforms. By contrast, both FOXM1b,

66 which contains none of the alternative exons, and FOXM1c which contains only the

Va exon, are transcriptionally active and can directly promote target gene expression in an isoform-specific manner (Ye et al., 1997). Interestingly, the FOXM1b isoform is expressed mostly in testis, suggesting that FOXM1 activity can be modulated through tissue-specific alternative splicing (Yao et al., 1997). The transcriptional activity of FOXM1 is regulated in part through dynamic interactions between its different domains. The full-length protein has three main components, including a central DNA-binding Forkhead box (FKH), an N-terminal auto-inhibitory domain

(NRD) and a C-terminal transcriptional transactivation domain (TAD) (Figure 1.9B).

Figure 1.9: The FOXM1 gene and its mRNA and protein structure. (A) The FOXM1 gene is composed of 10 exons. Va and VIIa can be spliced to generates three different splice variants, encoding for three FOXM1 protein isoforms, namely FOXM1a, FOXM1b, and FOXM1c. (B) The structure and phosphorylation sites of FOXM1. Phosphorylation by PlK1 increases FOXM1 activity, while CDK/CyclinA relieves the inhibitory function of the NRD. CHK2 enhances FOXM1 stability and Raf/MEK/MAPK mediated phosphorylation stimulates nuclear translocation. (NRD: N-terminal Repressor Domain, FKH: Forkhead DNA Binding Domain, TAD: Transactivation Domain, NLS: Nuclear localization signals)

1.8.2 FOXM1 in cell cycle and proliferation

In normal cells, the FOXM1 transcription factor plays an essential role in the regulation of a wide spectrum of cellular activities, including cell proliferation, cell cycle progression, cell differentiation, DNA damage repair, tissue homeostasis, and apoptosis (Koo et al., 2012). FOXM1 is considered as a critical cell cycle regulator, which controls the expression of genes required for both G1/S and G2/M transition

(Laoukili et al., 2005; Wang et al., 2005). It is also essential for mitotic entry and progression, ensuring the maintenance of chromosome stability (Laoukili et al.,

2005). The expression and the transcriptional activity of FOXM1 depend on the progression of cell cycle. FOXM1 mRNA and protein levels are actively high in proliferating cells at the entry to S phase and persist throughout the G2 and M- phases, and then gradually degrade during mitotic exit (Korver et al., 1997a, Laoukili et al., 2007). Conversely, FOXM1 expression is decreased in quiescent and terminally differentiated cells (Laoukili et al., 2008a, Yao et al., 1997, Korver et al.,

1997a). FOXM1 induces the expression of Cyclin A2, JNK1, ATF2 and CDC25A phosphatase, all of which are critical for G1/S transition and DNA replication (Wang et al., 2002). In addition, FOXM1 also regulates the transcription of Skp2 and Cks1, essential for the ubiquitinylation and degradation of the Cyclin-dependent kinase inhibitors (CKIs), p21CIP1 and p27KIP1, suggesting that FOXM1 negatively regulates the stability of p21CIP1 and p27KIP1, leading to increased CDK2 activity and enhanced

G1/S transition (Costa, 2005, Wang et al., 2005a, Wang et al., 2008).

Progression through the G2/M transition requires activation of the CDK1-Cyclin B complex through the removal of inhibitory phosphates at Thr-14 and Tyr-15 by the

CDC25B and CDC25C phosphatases (Borgne and Meijer, 1996, Wells et al., 1999,

Timofeev et al., 2010). Inactivation of FOXM1 results in diminished expression of

CDC25B and delayed accumulation of Cyclin B1, inhibiting CDK1/Cyclin B kinase activation and blocking progression into mitosis (Costa et al., 2005a). Moreover, during mitosis, FOXM1 also controls the expression of several genes critical for the spindle assembly checkpoint, chromosome segregation and cytokinesis, such as

Aurora B kinase, Suvivin, Polo-like kinase 1 (PLK1), CENP-A, B and F isoforms

(Laoukili et al., 2008a, Koo et al., Wang et al., 2005a). Consistently, inhibition of

FOXM1 expression leads to cell cycle arrest, mitotic spindle defects, chromosome mis-segregation, and mitotic catastrophe (McGovern et al., 2009). Collectively, these published studies indicate that FOXM1 is a crucial cell cycle regulator. It stimulates the expression of multiple proteins required for DNA replication and mitosis (Figure

1.10). In late M phase and early G1, FOXM1 is actively degraded during mitotic exit by the APC/C complex to prevent unscheduled expression of mitotic proteins and this degradation also requires Cdh1, a well-known co-factor of APC/C (Laoukili et al.,

2008a). Interestingly, a recent study in our laboratory has provided evidence that

FOXM1 activity is also controlled by SUMO1 conjugation and deconjugation.

SUMOylation of FOXM1 promotes its degradation in the cytoplasm by the APC/Cdh1 complex, leading to the silencing of FOXM1 activity and, consequently, mitotic delay

(Myatt et al., 2014). In addition, FOXM1 itself has an auto-repressive function. The interaction of N-terminal auto-inhibitory domain with the C-terminal half of the transcription factor leads to the inactivation of FOXM1 activity (Laoukili et al., 2008b,

Park et al., 2007). The phosphorylation of pRb by CDK4/Cyclin D1 has been shown to be important for relieving FOXM1 repression by disrupting the FOXM1 N-terminus

TAD interaction thereby strongly stimulating FOXM1 transcriptional activity for G1/S transition, whereas CDK2/Cyclin A complex is required to phosphorylate and activate

FOXM1 during G2 (Laoukili et al., 2008b, Wierstra and Alves, 2006).

Figure 1.10: FOXM1 is a crucial cell cycle regulator of genes essential for DNA replication, G2/M progression, chromosomal segregation and cytokinesis. FOXM1 is expressed in G1 phase and reaches its maximum level in late G1. This is maintained throughout G2 and mitosis. FOXM1 is translocated to the nucleus by the activation of MAPK signalling. FOXM1 transcriptional activity requires binding of the CDK2/Cyclin A during G1/S phase or CDK1/Cyclin B during G2/M phase. The binding of CDK1/Cyclin B complexes to its C-terminal transactivation domain is required for efficient phosphorylation of the CDK site, thereby mediating recruitment of the CBP coactivator protein. FOXM1 transcriptional target genes involved in G2/M progression include, PLK1, CENP-F, CDC25B, Aurora B and Cyclin B genes. In early G2 phase, PLK1 phosphorylates Cyclin B. This mediates its nuclear import and activates the onset of mitosis. PLK1 also participates in mitotic exit by activating the CDC20–APC/C complex to degrade the CDK1/Cyclin B complex. Aurora B kinase activity is required for both cytokinesis and accurate chromosome segregation because it regulates the localisation of several spindle assembly checkpoint proteins, such as BubR1 and Mad2. FOXM1 also regulates transcription of CENP-A and CENP-B, both of which are essential for kinetochore assembly. (Adapted from (Costa, 2005, Wang et al., 2005a))

1.8.3 FOXM1 in tumorigenesis and cancer progression

Since FOXM1 has been shown to play critical roles in regulating the expression of genes involved in cell proliferation and cell cycle progression, it is not surprising that elevated expression of FOXM1 has been detected in a wide range of cancer cell lines as well as in tumour tissues obtained from cancer patients (Koo et al., 2012).

Some of the early studies revealed that FOXM1 is upregulated in basal cell carcinoma (BCC) in comparison to normal skin cells (Teh et al., 2010). Genome- wide gene expression profiling analysis identified FOXM1 as one of the most commonly upregulated genes in solid tumours of the breast, prostate, lung, bladder, ovary, colon, liver, kidney, stomach, and pancreas (Pilarsky et al., 2004, Uddin et al.,

2011). An aberrant increase in the activity or expression of FOXM1 has also been found in the development of a variety of other cancers including cervical cancer

(Chan et al., 2008), esophageal squamous cell carcinoma (Hui et al., 2012), head and neck squamous cell carcinoma (Gemenetzidis et al., 2009), nasopharyngeal carcinoma (Chen et al., 2012), ovarian cancer (Zhao et al., 2014), acute myeloid leukaemia (AML) (Nakamura et al., 2010), oral cavity squamous cell carcinoma

(Chen et al., 2009) and malignant peripheral nerve sheath tumours (Yu et al., 2011).

More importantly, elevated expression of FOXM1 is also correlated with clinical aggressive behaviour and poor prognosis of numerous human cancers, suggesting that FOXM1 can serve as a novel prognostic marker for cancer patients (Yu et al.,

2011, Zhao et al., 2014, Yang et al., 2009, Koo et al., 2012). Therefore, it has become apparent that FOXM1 not only promotes tumorigenesis by increasing proliferative capacity and leading to uncontrolled cell division at the early stage of cancer development but also affecting several features of cancer progression including enhanced angiogenesis, tumour invasion, migration and metastasis,

71 increasing the resistance of cancer cells to chemotherapy, inducing replicative immortality, contributing to genomic instability and metabolism (Zhang et al., 2008,

Raychaudhuri and Park, 2011, Halasi and Gartel, 2013b, Koo et al., 2012, S.C.

Wilson et al., 2011).

The role of FOXM1 in angiogenesis is strongly associated with the activation of vascular endothelial growth factor (VEGF), which is the main angiogenesis protein produced and secreted by cancer cells. FOXM1 has been reported to directly regulate the promoter activity of VEGF. Depletion of FOXM1 caused decreased expression and activity of VEGF resulting in impaired angiogenesis, whereas overexpression of FOXM1 increased VEGF expression and promoted angiogenesis in brain, gastric, and breast cancer cell lines. This evidence suggests that FOXM1 is required for VEGF induced angiogenesis in cancer cells (Zhang et al., 2008, Li et al.,

2009, Karadedou et al., 2012, Koo et al., 2012). Moreover, FOXM1 has also been described as the master regulator of tumour metastasis, due to its effects on various aspects of the metastatic process, including epithelial-mesenchymal transition

(EMT), migration, invasion, and pre-metastatic niche formation (Halasi and Gartel,

2013a, Koo et al., 2012, Raychaudhuri and Park, 2011). Matrix metalloproteinases

(MMPs) have a critical role in the processes of tumour cell invasion and metastasis by degrading the basement membrane collagen (Overall and Kleifeld, 2006, Wang et al., 2010). It has been shown that downregulation of FOXM1 causes decreased expression and activity of MMP-2 and MMP-9, leading to reduced migration and invasion of pancreatic, breast, and bone cancer cell lines (Wang et al., 2007, Ahmad et al., 2010, Dai et al., 2007). FOXM1 directly activates the transcriptional activity of

MMP-2 (Dai et al., 2007), and indirectly regulates MMP-9 through its target JNK1 to induce migration and invasion of cancer cells (Wang et al., 2008). Moreover,

72 overexpression of FOXM1 with the loss of its inhibitor p19Arf induces liver cancer cells to undergo metastasis to the lung (Park et al., 2010). From the same study,

Park and colleagues also observed that FOXM1 directly regulates the expression and activity of Stathmin, lysyl oxidase (LOX), and lysyl oxidase like‐2 (LOXL2) to increase cell invasiveness in cancer cells (Park et al., 2010).

1.8.4 FOXM1 and DNA damage

Recently, the novel role of FOXM1 in DNA damage repair and in the maintenance of genome stability has been demonstrated in Foxm1-deficient MEF cells which exhibit defects in cytokinesis and chromosome mis-segregation as well as an increase in

DNA breaks (Tan et al., 2007). High levels of DNA damage, as evidenced by γH2AX foci, were also observed in U2OS cells depleted of FOXM1 using siRNA, and this correlated with reduced expression of the base excision repair factor X-ray cross- complementing group 1 (XRCC1) and breast cancer-associated gene 2 (BRCA2), two genes involved in DNA repair (Tan et al., 2007). Although both genes have been proposed as FOXM1 targets in U2OS cells, knockdown of FOXM1 in breast cancer cells did not cause the downregulation of its proposed downstream targets, XRCC1 and BRCA2 (Kwok et al., 2010b). This study suggested that FOXM1 might regulate the expression of other genes involved in DNA damage repair pathways (Kwok et al.,

2010b). Interestingly, a recent study from our laboratory has demonstrated for the first time that FOXM1 is required for DNA double strand break (DSB) repair by HR repair but dispensable for NHEJ repair (Monteiro et al., 2013). FOXM1 has also been identified as a direct transcriptional regulator of HR repair proteins, such as BRIP

(Monteiro et al., 2013), Rad51 (Zhang et al., 2012), BRCA2 (Tan et al., 2007), EXO1

(Zhou et al., 2014), RFC4, and cyclin D1 (Jirawatnotai et al., 2011). Interestingly,

FOXM1 not only controls DNA damage repair, but FOXM1 itself is also regulated in

73 response to DNA damage. Treatment with DNA damaging agents, such as IR and

DNA damaging agents has been reported to induce FOXM1 phosphorylation by

CHK2 , leading to its stabilisation at the protein level (Tan et al., 2007).

1.8.5 FOXM1 and Chemotherapy resistance

Besides mitosis, tumorigenesis and DNA Repair, FOXM1 has also been implicated in mediating chemotherapy sensitivity and resistance in various types of cancer including breast cancer (Francis et al., 2009). In breast cancer cell lines, upregulation of FOXM1 expression has been found to confer resistance to DNA damaging agents, such as cisplatin, doxorubicin and epirubicin, and its depletion was able to resensitise resistant cell lines to the respective cytotoxic drugs (Kwok et al., 2010a, Millour et al., 2011, Monteiro et al., 2013, Park et al., 2012). FOXM1 has also been related to genotoxic drug resistance in glioblastoma. FOXM1 knockdown sensitises resistant glioblastoma cells to alkylator temozolomide by downregulating the expression of DNA-repair gene Rad51 (Zhang et al., 2012). As DNA damaging agents exert their effect by inducing DNA damage and cell death and FOXM1 has been reported to drive the transcription of DNA repair genes. As a result, FOXM1 may promote resistance through enhancing the expression of DNA damage repair genes. Another possible epirubicin resistance mechanism through FOXM1 is mediated via the loss of p53 expression, which in turn suppresses E2F activity by the activation of the retinoblastoma pRB and/or the increase in ATM expression and activation (Millour et al., 2011). In addition, FOXM1 has also been associated with paclitaxel or docetaxel resistance in breast, ovarian, and gastric cancer cell lines

(Carr et al., 2010, Zhao et al., 2014, Li et al., 2014). Car and colleagues revealed that FOXM1 can alter microtubule dynamics by regulating directly the expression of the tubulin destabilising protein Stathmin in order to protect tumour cells from

74 paclitaxel-induced apoptosis (Carr et al., 2010). Therefore, targeting FOXM1 could be a viable strategy in circumventing acquired resistance.

Figure 1.11: Functional roles of FOXM1. FOXM1 regulates a variety of crucial cellular activities through regulating the transcriptional activity of its target genes, which is critical for cell cycle progression, cell proliferation, DNA damage repair, angiogenesis, cell migration, cellular senescence and chemotherapeutic drug resistance. Adapted from (Koo et al., 2012).

1.9 Thesis Aims

Epirubicin and paclitaxel are the most effective and widely used chemotherapeutic agents for treating patients with advanced or metastatic breast cancer, triple negative breast cancers or ER-positive cancers that have become resistance to endocrine therapy. Resistance to chemotherapy represents the main obstacle against the successful management of breast cancer. Understanding the molecular mechanisms underlying epirubicin and paclitaxel resistance is thus crucial for the development of effective chemotherapeutic strategies for breast cancer treatment.

Overexpression of the oncogenic transcription factor FOXM1 has been implicated in the development of drug resistance to various cytotoxic agents. However, the exact molecular mechanism underlying sensitivity and resistance to epirubicin and paclitaxel are not completely understood. Therefore, the aim of this study is to elucidate the role of FOXM1 in the development of epirubicin and paclitaxel resistance in breast cancer. Epirubicin and paclitaxel are known to act through distinct mechanisms. Therefore, this study has been divided into two main parts:

i. To investigate the role of FOXM1 in DNA damage response and epirubicin

resistance.

ii. To investigate the role of FOXM1 in mitotic control and paclitaxel resistance.

CHAPTER 2

MATERIALS AND METHODS

2.1 Cell culture and cell lines

2.1.1 The human breast cancer cell lines: MCF-7 and MDA-MB-231

The human breast cancer MCF-7 and MDA-MB-231 cell lines originated from the

American Type Culture Collection and were acquired from Cancer Research UK

(London, UK), where they were tested and authenticated. Cells were cultured in

Dulbecco’s Modified Eagle’s Medium (Sigma, Poole, UK) supplemented with 4 mM

L-glutamine and penicillin–streptomycin (Gibco), and 10% fetal calf serum (FCS)

(First Link Ltd, Birmingham, UK) in a humidified incubator at 37°C with 10% CO2.

2.1.2 MCF-7 EpiR and MCF-7 TaxR

The epirubicin resistant MCF-7 (MCF-7 EpiR) cells and the paclitaxel resistant MCF-7

(MCF-7 TaxR) cells were previously established in the laboratory by chronic exposure of the parental MCF-7 cells to stepwise increase concentrations of both drug until they acquired resistance to 10 µM epirubicin and 0.05 µM paclitaxel, respectively (Millour et al., 2011, Khongkow et al., 2013). Cells were cultured in

Dulbecco’s Modified Eagle’s Medium (Sigma, Poole, UK) supplemented with 4mM L- glutamine and penicillin–streptomycin (Gibco), and 10% fetal calf serum (FCS) (First

Link Ltd, Birmingham, UK) in a humidified incubator at 37°C with 10% CO2.

2.1.3 Mouse embryonic fibroblasts (MEFs)

The wild-type and Foxm1-/- mouse embryonic fibroblasts (MEFs) as previously described (Laoukili etal., 2005), were kind gifts from Professor René Medema

(Netherlands Cancer Institute, Amsterdam, the Netherlands). Cells were also cultured in Dulbecco’s Modified Eagle’s Medium (Sigma, Poole, UK) supplemented with 4mM L-glutamine and penicillin–streptomycin (Gibco), and 10% fetal calf serum

(FCS) (First Link Ltd, Birmingham, UK) in a humidified incubator at 37°C with 10%

CO2.

2.1.3 HeLa cells carrying DR-GFP

The HeLa cell line carrying the DR-GFP (direct repeat-green flouorescent protein) reporter construct, as described previously (Pierce et al., 1999), was a gift from

Professor Maria Jasin (Memorial Sloan-Kettering Cancer Center, New York, USA).

Cells were cultured as described above for MCF-7 cells.

2.1.4 Maintaining Cultured Cells

Adherent cells were cultured and split at approximately 80% confluency (80% of the flask surface covered by cell monolayer). To sub-culture, the culture medium was aspirated from the flask and the cells were gently washed once with sterile PBS

(Oxord Ltd, Basingstoke, UK) to get rid of any FCS in the residual culture media.

Cells were detached by incubation 1x trypsin diluted in EDTA (Sigma-Aldrich, UK) at

37°C. After being detached, the cells were then resuspended in the completed media to stop the trypsin reaction and transferred into the new flask or the appropriate cell culture dishes.

2.1.5 Cell line storage

Healthy and low passage cells were trypsinised, as described above, resuspended in culture media and transferred into a sterile centrifuge tube. The cells were collected by centrifuging at 800×g for 4 min. The cell pellet was resuspended in 1 mL of sterile freezing solution, containing 10% di-methyl sulfoxide (DMSO) in FCS (Sigma-

Aldrich, UK) and then transferred into suitable vials or cryotubes. The tube was placed into the isopropanol freezing box (VWR International, Lutterworth, UK) at -

80°C overnight before storage in liquid nitrogen.

For defrosting cells, cryotubes were rapidly thawed by placing them in a 37°C water bath. Cell suspension was mixed with 1 mL of complete medium and centrifuged at

800×g for 4 min. The freezing solution was discarded and the cell pellet were resuspended in the fresh complete medium and placed into an appropriate culture flask to recover and grow.

2.2 Drug treatment

Epirubicin (2 mg/mL) (Medac, Germany) and Paclitaxel (6 mg/mL) (Teva UK Limited,

East Sussex, UK) were obtained from Imperial College Healthcare, UK. Both chemotherapeutic agents were stored at 4°C. Thiostrepton (from Streptomyces azureus, 90% purity) was obtained from Sigma, UK and stored at -20°C.

2.3 Plasmid constructs

The pcDNA3-FOXM1B expression plasmid has been described previously (Leung et al., 2001, Kwok et al., 2010). The pmCherry-FOXM1B expression plasmid was generated by cloning the full-length FOXM1 cDNA from pcDNA3-FOXM1B into the

EcoRI and BamHI sites of the pmCherry-N1 vector (Clontech, Mountain View, CA,

USA). The pCMV-I-Scel was a gift from Dr. Joanna Morris (University of

Birmingham, UK). The human NBS1 promoter-luciferase constructs have previously been described (Chiang et al. 2003). Putative forkhead site mutagenesis within the

NBS1 promoter was generated in the laboratory using Quickchange Site-Directed

Mutagenesis Kit according to the manufacturer’s instructions (Agilent technologies,

CA, USA).

2.3.1 Cloning of WT and mut KIF20A luciferase reporter constructs

To generate KIF20A-WT and KIF20A-MUT luciferase reporter constructs, the

KIF20A-WT, KIF20A-MUT1, and KIF20A-MUT2 fragments (Figure S10) were initially designed and synthesised by GeneArt (Life Technologies, Darmstadt, Germany).

The synthesised DNA fragments were then amplified and cleaved with the restriction enzymes XhoI and Bglll and cloned into the XhoI/Bglll sites of the pGL3-basic vector

(Promega, Madison, WI). DNA extracted from all the positive clones were re- digested with XhoI and Bglll enzymes and performed gel electrophoresis to confirm the gene inserts (as shown in the Figure 4.7B).

2.4 Transfection

2.4.1 Gene Silencing with Small Interfering RNAs (siRNAs)

For gene silencing, cells were transiently transfected with siRNA SMARTpool reagents purchased from Thermo Scientific Dharmacon (Lafayette, CO, USA) using

Oligofectamine (Invitrogen, UK) according to the manufacturer’s instructions. All siRNAs used for the work in this thesis were ON-TARGETplus SMARTPool siRNAs

(Dharmacon Thermo Scientific, CO, USA). This technology comprises a mixture of 4 distinct siRNAs against the target gene. Pooling reduces the concentration of each individual siRNA therefore decreasing potential off-target effects. The SMARTPool siRNAs used were: FOXM1 siRNA (L-009762-00) targeting all isoforms of FOXM1,

NBS1 siRNA (L-009641-00), KIF20A siRNA (J-004957-06) and the NS control siRNA

(D-001810-10-05), confirmed to have minimal targeting of known genes. All the siRNA were diluted to 20 µM in 1X siRNA buffer and stored at -20°C. Briefly, one day before transfection, cells were plated into six-well culture dishes to get a 70%-80% confluency at the time of transfection. For each well, 5 µL Oligofectamine (Invitrogen,

UK) was diluted with 60 µL of Opti-MEM (Life Technology, UK), mixed gently and incubate at room temperature for 5 min. Then the siRNA dilution was prepared by adding 7.5 µL of 20 µM siRNA stock in 250 µL of Opti-MEM. The diluted

Oligofectamine was combined with the diluted siRNA, vortexed briefly and incubated for 20 min at room temperature to allow the siRNA/Oligofectamine complexes to form. Opti-MEM was added to final volume of 500 µL. The culture media was removed and the cells were washed once with warm PBS, the siRNA-Oligofectamine transfection mixtures were then added to the cells. The plate was returned to incubate at 37°C. 6 h after transfection, the completed media was added to each well. The cells were incubated at 37°C in a CO2 incubator for 24-72 h until they are ready to assay for gene knockdown.

2.4.2 FuGENE 6 transfection

The cells were transfected with DNA plasmids using the FuGENE 6 transfection reagent (Roche Diagnostics, West Sussex, UK) at a ratio of 1:3 ratio (1 µg of DNA plasmid: 3 µL of FuGENE 6). Briefly, one day before transfection, cells were plated into six-well culture dishes to get a 50%-60% confluent at the time of transfection.

FuGENE 6 was diluted in serum-free DMEM medium (Sigma-Aldrich, Poole, UK) and incubated for 5 min at room temperature. 1 µg of the DNA plasmids was added into the diluted FuGENE 6, vortexed briefly and incubated for 15 min at room temperature to allow the DNA/Fugene 6 complex to form. The DNA/Fugene 6 complex was then added into the cells in a drop-wise manner. Cells were then incubated at 37°C in a CO2 incubator for 24-72 h until they were ready to be assayed for gene expression.

2.4.3 Xfect mESC transfection

Xfect Mouse Embryonic Stem Cell Transfection Reagent (Xfect mESC) was used to transfect MEFs at a 2:1 ratio (5 µg of DNA plasmid: 2.5 µL of Xfect mESC reagent).

One day before transfection, early passage of WT or Foxm1-/- MEFs were plated into six-well culture dishes to get an approximately 70-80% confluent at the time of transfection. For each transfection, 5 µg of plasmid DNA was diluted in 95 µL of

Xfect reaction buffer, and 2.5 µL of Xfect mESC polymer was diluted separately in

97.5 µL of Xfect reaction buffer. The diluted plasmid DNA was then mixed with the

Xfect mESC polymer dilution and incubated at room temperature for 10 min to allow nanoparticle complexes to form. The nanoparticle complex solution was next dropwised to the cells and incubated at 37°C in a CO2 incubator for 3 h before replacing with 2 mL fresh media. Cells were then returned to the 37°C incubator for

48-72 h until they were ready to be assayed for gene expression.

2.4.4 Co-transfection of plasmid DNA and siRNA using Lipofectamine 2000

HeLa cells containing DR-GFP construct were co-transfected with plasmid DNA and siRNA using the Lipofectamine2000 transfection reagent (Life technologies, UK).

One day before transfection, cells were plated into six-well culture dishes to get a 70-

80% confluent at the time of transfection. For each well, 5 µL Lipofectamine 2000

(Invitrogen, USA) was diluted with 244 µL of Opti-MEM (Life Technologies, UK) mixed gently and incubate at room temperature for 5 min. Then the DNA plasmid/siRNA dilution was prepared by adding 2.5 µL of 20 µM siRNA stock and 7.5

µL of 100ng/uL DNA plasmid in 240 µL of Opti-MEM. The diluted Lipofectamine was combined with the diluted DNA plasmid-siRNA, vortexed briefly and incubated for 20 min at room temperature to allow the DNA-siRNA/Lipofectamine complexes to form.

The DNA-siRNA/Lipofectamine transfection mixture was then added into the cells in a drop-wise manner and incubated at 37°C in a CO2 incubator for 6 h before replacing with 2 mL fresh media. Cells were then returned to the 37°C incubator for

48-72 h until they were ready to be assayed for gene expression.

2.5 Quantitative real time PCR

2.5.1 RNA extraction and quantification

Total RNA was isolated from cell pellets using the RNeasy kit (Qiagen, Crawley, UK) according to manufacturer’s instructions. The RNA purity and concentration were then determined by measuring the spectrophotometric absorbance at 260 nm and

280 nm with a ND-1000 NanoDrop (Thermo Fisher Scientific, Wilmington, Delaware

USA).

2.5.2 cDNA synthesis (Reverse transcription)

The RNA was reverse transcribed into first strand cDNA using Superscript III First-

Strand system (Life technologies, UK). Briefly, 2 µg of RNA was mixed with 1 µL of

50 µM Oligo (dT)20 primer, 1 µL of 10mM dNTPs mix and DEPC-treated water up to

13 µL. Samples were denatured at 65°C for 5 min and then place on ice for 1 min.

Subsequently, 6 µL of cDNA synthesis solution (consisting of 1 µL of the reverse transcriptase superscript III (200U/µl), 1 µL of 0.1 M DTT, 1 µL of RNaseOUT recombinant inhibitor and 4 µL of 1X first strand buffer) was added into the denatured RNA, mix gently and briefly centrifuged to collect the liquid down at the bottom of the tube. All the cDNA synthesis solution’s components were obtained from Life technologies. The reaction tube was placed into a thermocycler and run the following programme: incubated at 25°C 10 min for the initial step, then 50 min at

50°C for the reverse transcription step, and at 70°C for 15 min to inactivate the reverse transcription enzyme. The cDNA mixture was diluted 1:4 with DEPC-treated water and used immediately for RT-qPCR or stored at -20°C.

2.5.3 Primers Design and optimisation

All primers (Table 2.1 and 2.2) were designed using Primer Express 3 software

(Applied Biosystems) and purchased from ABI (Life technologies).

The forward and reverse human primers used were:

Target gene 5’-Forward-3’ 5’-Reverse-3’

FOXM1

(detection of TGCAGCTAGGGATGTGAATCTTC GGAGCCCAGTCCATCAGAACT all isoforms)

NBS1 TTTTCAACCAGTTTTCCGTTACTTC ACACTGCGCGTATAAGCCAAT

ATM AATATCCATTCACCGCAGCC TGCCCATCCGGGACAA

MRE11 TGAGAACTGGCCTTCGATTCA GGAGCCCAGACAAGCATGAT

RAD50 TCCAAATCTTGTGGAAGTGCAT CTGCAAGCAGCCAGAACTTG

KIF20A GCCAACTTCATCCAACACCT GTGGACAGCTCCTCCTCTTG

L19 GCGGAAGGGTACAGCCAAT GCAGCCGGCGCAAA

Table 2.1: List of human RT-qPCR primers used in this study

The forward and reverse mouse primers used were:

Target 5’-Forward-3’ 5’-Reverse-3’

Foxm1 AGAAATGTGACCATCAAAACTGAAAT GAGGGAGCAGAGGCTTCATCTT

Nbs1 TGACAACCCGATAGAGGAGCAT TCTTGGCTCTCTGTCTGTCCAG

Atm GCGCCACGCCTTGT CAAACGTTGCCTGAAT

L19 CCCGTCAGCAGATCAGGAA GTCACAGGCTTGCGGATGA

Table 2.2: List of mouse RT-qPCR primers used in this study

2.5.4 Real time quantitative PCR (RT-qPCR)

RT-qPCR analysis was carried out on a 7900HT Fast Real-time PCR System

(Applied BioSystems, UK) using Power SYBR green (Applied Biosystem, UK). SYBR

Green is a DNA binding dye that intercalates tightly to double-stranded DNA. This intercalation causes the SYBR to fluoresce. The qPCR machine detects the fluorescence and software calculates Ct values from the intensity of the fluorescence. Transcript levels were quantified using the relative standard curve method. All analyses were carried out in triplicate, and the target gene expression levels were normalised to a control mRNA level of ribosomal protein L19.

Briefly, a mixture containing optimised concentration of gene-specific forward and reverse primer pairs with 12 µL of SYBR-Green Master Mix was added to 2 µL of the cDNA (~100ng). For each reaction, DEPC-treated water was added to adjust the final volume to 25 µL. Transcript levels were quantified using the relative standard curve method. The standard curve was made by mixing 5 µL from each cDNA sample into a single tube and diluted 1:4 for four serial dilutions. The standard dilution included 1, 1:4, 1:16, 1:64, 1:256, and NT (only water, no cDNA). All analyses were carried out in triplicate. The PCR program was set to 95°C for 10 min for the enzyme activation, followed by 40 cycles of denaturing at 95°C for 10 sec, and primer annealing at 55°C for 30 sec. After the real-time PCR run, a dissociation curve was generated to confirm the absence of non-specific amplification.

2.6 Western Blot Analysis

Western blot analysis was used to detect protein expression. A complex mixture of proteins extracted from the cells was separated according to molecular weight using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Following electrophoresis, the separated proteins were transferred into nitrocellulose membrane for immunoblotting.

2.6.1 Protein extraction and quantification

Cells were harvested by trypsinisation as described above. Cell pellets were washed with PBS, transferred into a new eppendof tube, and centrifuged at 300×g for 4 min.

To prepare whole cell extracts, cell pellets were lysed in lysis buffer, which consists of 1% (v/v) Nonidet P-40, 50 mM Tris-HCL (pH7.4), 0.5 mM EDTA, 150 mM NaCL,

50 mM HEPES, 1 mM sodium fluoride (NaF), 1 mM sodium orthovanadate (Na3VO4),

2 mM phenylmethysulfonyl fluoride (PMSF) and complete protease inhibitors mixture

(Roche Applied Science), and incubated on ice for 15 min. The protein extracts were then centrifuged at 14,000 rpm for 10 min at 4 °C to remove insoluble cellular debris.

The supernatant was then transferred in to the new eppendof tube and protein quantification was determined using Bio-Rad Dc protein assay according to manufacturer's protocol. Absorbance of the samples was read at 700 nm using the

Sunrise spectrophotometer (Tecan, Reading, UK).

2.6.2 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

SDS polyacrylamide gel electrophoresis (SDS-PAGE) involves the separation of

proteins based on their molecular size. The standard SDS-PAGE gels composed of

two different layers, an upper stacking gel and a lower running gel. The gel was

made with varying amounts of bis-acrylamide solution, the percentages of the gel

used varying for 5 to 14. Lower percentage of gels allowed better examination of

proteins of higher molecular weight. The bis-acrylamide with ammonium persulphate

(APS) and tetraethylethyleneduanine (TEMED) were the catalysts for gel

polymerisation. Table 2.3 showed the components of different percentages of gels.

Resolving gel Stacking gel Reagents 5% 7% 10% 12% 14% 5% dH20 (mL) 6.19 5.02 4.02 3.35 2.68 3.67 1.5M Tris pH 8.8 (mL) 2.5 2.5 2.5 2.5 2.5 - 1.5M Tris pH 6.8 (mL) - - - - - 0.42 30% Acrylamide/0.8% Bis (37.5:1) (mL) 1.16 2.33 3.33 4.00 4.67 0.83 10% SDS (µL) 100 100 100 100 100 50 25% APS (µL) 40 40 40 40 40 20 TEMED (µL) 10 10 10 10 10 10 Total (mL) 10 10 10 10 10 5 Table 2.3: Components of the difference percentage of SDS-PAGE gels

For protein expression analysis, 20 µg of protein extracts were mixed with 6X SDS

protein dye (0.375 M Tris pH 6.8, 12% SDS, 60% glycerol, 0.6 M DTT, 0.06%

bromophenol blue), boiled at 100°C for 10 min. Samples were then loaded into the

SDS-PAGE gels in Tris-glycine electrophoresis running buffer (25 mM Tris-Base,

109 mM Glycine, and 0.1% (w/v) SDS) and run at a constant voltage of 90 volts.

After electrophoresis, proteins were electro-transferred onto nitrocellulose

membranes (Whatman International, Kent, UK) using Bio-Rad Mini Trans-Blot

system, filled with transfer buffer ( 25M Tris-Base, 190 mM glycine and 20% (v/v)

ethanol) and run at a constant 90 volts for 90 min at room temperature.

2.6.3 Immunoblotting

The membranes were blocked with 5% bovine serum albumin (BSA) for 30 min at room temperature and then incubated with specific primary antibodies overnight at

4°C. Next day, the membranes were washed three times, 10 min each with TBS-T.

Secondary antibodies (DAKO, Ely, UK) diluted 1:2000 in TBS-T were added for 45 min at room temperature and membranes were washed three times with TBS-T.

Immunoblots were visualised using the ECL detection system (Amersham

Biosciences, UK).

2.6.4 Antibodies

Primary antibodies used in this study are shown in the Table 2.4. Secondary antibodies used were goat anti-mouse and goat anti-rabbit HRP (horseradish peroxidase) were from Dako (Ely, UK). For immunostaining studies, Alexa Fluor 488 goat anti-rabbit IgG (H+L), Alexa Fluor 488 goat anti-mouse IgG (H+L), Alexa Fluor

555 goat anti-rabbit IgG (H+L), and Alexa Fluor 555 goat anti-mouse IgG (H+L) were obtained from Molecular Probes (Invitrogen, Carlsbad, CA, USA).

NO. Antibody Product number Species Company Dilution

1 ATM 04-200 Rabbit Calbiochem 1:1000

2 CHK2 2662 Mouse Cell Signaling 1:1000 FOXM1 3 (detection of all C-20/sc-502 Rabbit Santa Cruz 1:1000 isoforms) 4 H2AX 2595 Rabbit Cell Signaling 1:1000

5 I-Sce1 sc-98269 Rabbit Santa Cruz 1:1000

6 MRE11 NB-100-142 Rabbit Novus Biologicals 1:1000

7 NBS1 3002 Rabbit Cell Signaling 1:1000

8 PARP 9542 Rabbit Cell Signaling 1:1000

9 p-ATM (Ser1981) MAB3806 Mouse Millipore 1:500

10 p-CHK2 (Thr68) 2661 Rabbit Cell Signaling 1:1000

11 p-NBS1 (Ser343) 3001 Rabbit Cell Signaling 1:1000

12 RAD50 NB100-154 Mouse Novus Biologicals 1:1000

13 β-tubulin (H-235) H-235/sc-9104 Rabbit Santa Cruz 1:1000

14 γ-H2AX (Ser 139) 9718 Rabbit Cell Signaling 1:1000

15 Cyclin B1 H-433/ sc-752 Rabbit Santa Cruz 1:1000

16 KIF20A Ab104118 Rabbit Abcam 1:1000

17 KIF20A D-3/ sc-374508 Mouse Santa Cruz 1:1000

18 Cleaved-caspase7 9491 Rabbit Cell Signaling 1:1000

19 p21 sc-397 Rabbit Santa Cruz 1:1000

20 p53 sc-374087 Mouse Santa Cruz 1:1000

21 Rabbit HRP P0448 goat Dako 1:2000

22 Mouse HRP P0447 goat Dako 1:2000 Table2.4: List of primary antibodies used for western blotting

2.7 Immunofluorescent Staining

2.7.1 γ-H2AX assay

Cells grown on chamber culture slides were fixed in 4% paraformaldehyde (Thermo

Scientific, Rockford, IL, USA) for 15 min followed by permeabilisation for 10 min with

0.2% Triton X-100 in PBS. After being washed with PBS, cells were blocked with 5% goat serum in PBS for 30 min. Slides were incubated overnight at 4°C with rabbit polyclonal anti-γ-H2AX pS1981 primary antibody (Cell signalling, UK) diluted 1:250 in 0.2% goat serum, rinsed with PBS and subsequently incubated with a 1:500 dilution of Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody

(Molecular Probes, Invitrogen) for 45 min at room temperature. After washing with

PBS, cells were stained with DAPI for 10 min and mounted with Vectashield mounting solution (Vector Laboratories). Images were captured with a Leica TCS

SP5 confocal microscope (Leica Microsystems, Mannhein, Germany) equipped with a 63X oil immersion objective and LAS-AF software. For each experimental condition, at least 100 cells were captured and quantified for the number of γ-H2AX foci using FociCounter software.

2.7.2 α-tubulin immunostaining to detect mitotic catastrophe

Cells grown on chamber culture slides were fixed in 4% paraformaldehyde (Thermo

Scientific, Rockford, IL, USA) for 15 min followed by permeabilisation for 10 min with

0.2% Triton X-100 in PBS. After being washed with PBS, cells were blocked with 5% goat serum in PBS for 30 min. The slides were incubated overnight at 4 °C with primary antibodies, anti-α-tubulin (clone DM1A) and anti-γ-tubulin purchased from

Sigma-Aldrich (St. Louis, MO, USA). The slides were then rinsed with PBS and subsequently incubated with a 1:500 dilution of Alexa Fluor 488-conjugated goat anti-mouse and Alexa Fluor 555-conjugated goat anti-rabbit secondary antibodies

(Molecular Probes, Invitrogen) for 45 min at room temperature. After washing with

PBS, cells were mounted with Vectashield mounting solution containing DAPI

(Vector Laboratories). Mitotic cells were visualised with a Leica TCS SP5 confocal microscope (Leica Microsystems, Mannhein, Germany) equipped with a 63X oil immersion objective and LAS-AF software. For each condition, images of at least 50 mitotic cells were captured and analysed.

2.8 Luciferase reporter assay

Cells were seeded into 96-well plates at a density of approximately 10,000 cells/well and co-transfected with a firefly luciferase reporter plasmid, the transfection control

Renilla plasmid (pRL-TK; Promega, Southampton, UK) and pcDNA3-FOXM1 plasmids using FuGENE 6 transfection reagent (Roche Diagnostics, UK). Twenty- four hours after transfections, cells were harvested to determine luciferase activity, using the Steadylite plus™ Reporter Gene Assay (Perkin Elmer, UK) according to manufacturer’s instruction. Luminescence was then determined using a PHERAstar

Plus microplate reader (BMG Labtech, Aylesbury, UK). Each firefly luciferase was normalised by well to the Renilla readings and six replicates were measured per condition.

2.9 HR Repair Assay

HeLa cells carrying DR–GFP reporter system were used to measure HR repair as described (Morris et al., 2009, Pierce et al., 1999). Cells were transfected with either siRNA or DNA expressing plasmid. Cells were left for 24 h before transfection with pCMV- I-Scel plasmid to induce DSBs and stimulates HR repair. 72 h after I-Scel transfection, cells were harvested and the number of GFP positive cells was determined by BD FACS Canto II flow cytometer system (BD Biosciences, Oxford,

UK) using the FACSDiva acquisition software (BD Bioscience, UK). The result were analysed by using a two-dimensional dot plot (green fluorescence (FL1-H) on y-axis and auto-orange florescence on the x-axis (FL2-H). For each condition, 50,000 cells were analysed and the frequency of HR events was calculated from the number of

GFP-positive cells divided by the total number of cells analysed.

2.10 Sulforhodamine B (SRB) assay

The sulforhodamine B (SRB) assay was used to measure drug-induced cytotoxicity and cell proliferation. The assay relies on the ability of SRB, a bright-pink aminoxanthene dye, to bind with basic amino-acid residues in protein components of cells under acidic conditions. As the binding of SRB is stoichiometric, the amount of dye extracted from stained cells is directly proportional to the cell mass (Vichai and

Kirtikara, 2006). Approximately 3,000 cells were plated in each well of the 96-well plates (Falcon, USA). 24 h after culturing, cells were treated with different concentrations of epirubicin or paclitaxel for 48 h. After an incubation period, monolayer cells were fixed with 100 µL of 40% (w/v) trichloroacetic acid (TCA) and incubated for 1 h at 4°C. The plates were then rinsed with water for three times before staining with 0.4% SRB solution (0.4% SRB in 1% acetic acid) for 1 h at room temperature. The excess dye is removed by washing repeatedly with 1% (v/v) acetic acid and the plated were dried overnight. The protein-bound dye was dissolved in 10

M Tris base solution for 30 min at room temperature with agitation. Absorbance was measured at 492 nm using a Sunrise spectrophotometer (Tecan, Medford, MA,

USA). The results were normalised to 0 h or untreated wells. All the reagents were obtained from Sigma-Aldrich.

2.11 Clonogenic Assay

Clonogenic assay, a useful long-term survival assay, was used to examine the effectiveness of anticancer cytotoxic agents on colony forming ability. Cells were seeded in 6-well plates at a density of approximately 2,000 cells/ well overnight. The cells were then treated with different concentrations of epirubicin or paclitaxel for 48-

72 h. The drug was removed and replaced with fresh completed media. The surviving cells were left to form colonies. After 14 days of incubation, cell colonies were fixed with 4% paraformaldehyde for 15 min at room temperature then washed with PBS for three times. 0.5% crystal violet was used to stain the fixed cells for 30 min. The excess crystal violet was removed and the plates were washed with tap water. Plates with colonies were then left to dry overnight. Digital images of the colonies (at least 5 random fields) were obtained using a camera. Quantification was achieved by solubilising dye with 1 mL of 33% (w/v) acetic acid and the absorbance was measured at 592 nm using a microplate reader (Tecan, Medford, MA, USA).

2.12 Senescence-Associated β-Galactosidase assay

Senescence-associated β-galactosidase (SA-β-gal), a widely used biomarker of senescence (Itahana et al., 2007), was detected by histochemical staining of cells using the artificial substrate X-gal. Briefly, cells were seeded 6-well plates at a density of approximately 20,000 cells/well before treatment with epirubicin or paclitaxel for 48 h. After culture for a further 5 days, cells were fixed and stained using a Senescence β-Galactosidase Staining Kit (#9860), purchased from Cell

Signalling Technology (Beverley, MA, USA). Plates were incubated overnight at

37°C in a dry incubator (no CO2). Cells were then detected for blue staining under a bright-field microscope. The percentage of SAβ-gal-positive cells was calculated by counting the cells in five random fields.

2.13 Cell Cycle Analysis

Cell cycle analysis was done by propidium iodide staining. Briefly, cell pellets were harvested, washed in 500 µl PBS, and spun at 300xg (4ºC) for 2 min before being fixed with 5 mL of 90% ethanol at 4ºC for overnight. Cells were then spun at 300xg

(4ºC) for 5 min. The supernatant was discarded and the cells washed twice in 2 mL

PBS. The supernatant was discarded and the cells were stained using 600 µl propidium iodide (50 µmol/l) for 1 h in the dark. The cell cycle profile was then analysed using a FACS canto and FACSDiva software (BD Biosciences).

2.14 Chromatin Immunoprecipitation (ChIP)

Cells were cross-linked in 1% formaldehyde for 10 min and glycine treated for 5 min before cell scraping and centrifugation. The pellets were washed with PBS+1X PI and resuspended in 1 mL LB1 buffer (50 mM HEPES-KOH, pH 7.5; 140 mM NaCl; 1

M EDTA; 10% glycerol; 0.5% Igepal CA-630; 0.25% TritonX-100+1X PI) for 10 min on a rotating platform at 4°C. Samples were centrifuged at 2000×g for 5 min at 4°C before lysis in 1 mL LB2 buffer (10 M Tris–HCl, pH 8; 200 M NaCl; 1 mM EDTA; 0.5 mM EGTA+1X PI) for 5 min on a rotating platform. Lysates were centrifuged as before and pellets were resuspended in 300 µl LB3 buffer (10M Tris-HCL, pH 8.0;

100 M NaCl; 1 M EDTA; 0.5 M EGTA; 0.1% Na-deoxycholate and 0.5% N- lauroylsarcosine+1X PI). DNA was fragmented to an average size of 150-200 bp using Bioruptor (Diagenode, Denville, USA) with the setting 4 cycles with 5 min duration. After sonication, TritonX-100 was added to a concentration of 1%. The mixture was centrifuged at 13,000×g for 10 min at 4°C. Five percent of each sample was taken as input and frozen until the next day. 40 µl per sample of Dynabeads conjugated to Protein A (Invitrogen) was transferred into fresh tubes and blocked by washing three times with PBS containing 0.5% bovine serum albumin (PBS-BSA) before incubation with 4 μg of the indicated antibodies overnight at 4°C. Anti-FOXM1 antibody (SC502) was from Santa Cruz Biotechnology (Santa Cruz, Heidelberg,

Germany) and rabbit IgG control (X0903) was obtained from Dako (Cambridgeshire,

UK). Dynabeads were then washed twice with PBS-BSA to remove unbound antibody, and the 100 µL chromatin samples prepared the previous day were added to the appropriate Dynabead samples and rotated overnight at 4°C. Dynabeads were subsequently washed six times in RIPA buffer (50 mM HEPES-KOH, pH 7.5; 0.5 M

LiCl; 1 mM EDTA; 1% NP-40; 0.7% Na-deoxycholate) and two times in TE buffer (20

98 mM Tris–HCl, pH 7.6; 150 mM NaCl) before incubating overnight in 100 μl elution buffer (1% SDS, 0.1 M NaHCO3) at 65°C to elute protein–DNA complexes and to reverse formaldehyde-induced cross-links. At the same time, input samples were defrosted and subject to elution. The next day, the supernatant was transfer to a fresh tube and diluted with 200 μl TE buffer. To get rid of contaminating RNA and protein, samples were incubated with 8 μl of 1 mg/mL RNAse and 4 μl 20 mg/mL proteinase K (Invitrogen), respectively. ChIP DNA was purified by phenol chloroform extraction and resuspended in deionised water and subject to quantitative real-time

PCR using primers specific to KIF20A gene (Table 2.5). Data were presented as %

Input using the following formula: % Input = 100 × 2^(CT adjusted Input sample − CT immunoprecipitated sample). At least two individual repeats of the entire experiment were conducted (** P<0.001).

Primer Forward (5’-3’) Reverse (5’-3’)

KIF20A TTCCTTACGCGGATTGGTAG AGCCGCAGAGCACAACTC

Negative control CCGCCTCCCTCTTAGCATAA CAGGAAATTGCATCTCGGGG

Table 2.5: Primers used for quantitative real-time PCR following ChIPs The negative control primer specifically binds and amplifies the region located at the 5’ upstream non- coding region (-813/-938 bp) of KIF20A gene.

2.15 Clinical data analysis

2.15.1 Tissue Microarray

A set of patients samples described in Chen et al., 2010 was used. Briefly, 116 cases of breast cancer diagnosed between the years 1992 to 2001 together with clinical follow up data were retrieved from the records of the Department of

Pathology, Queen Mary Hospital of Hong Kong. Histological features of all cases were determined by the pathologist, the representative paraffin tumour blocks were chosen as donor block for each case and the selected areas were marked for construction of tissue microarray (TMA) blocks. A total of 116 was assessed and scored for FOXM1, NBS1 and KIF20A staining. The expression pattern and subcellular localisation were analysed and correlated with various clinicopathological features, including histological type, histological grade, clinical stage, estrogen and progestrogen receptor status, lymph node metastasis as well as survival time.

2.15.2 Immnohistochemistry

The TMA slide was deparaffinised with xylene, rehydrated with decreasing concentrations of ethanol. Citrate buffer (0.01 M, pH 6.0) was used for antigen retrieval. To quench endogenous peroxidase, the slides were then immersed into 3%

H2O2 methanol for 10 min at room temperature. Following with rinsing in 0.05%

Tween in PBS (PBS-T) twice, 200 µL of 1:50 primary specific antibody dilution (Table

2.6) was added to each section and incubated at 4°C overnight. The slides were then washed in PBS-T and 4-5 drops of DAKO Polymer were then added onto each sample and incubated at room temperature for 30 min. After washing, Chromogen

DAB/substrate reagent was applied on the slides and incubated for 3 min. Finally,

100 the dehydration of slides was performed using increasing concentrations of ethanol solution, followed by clearing in xylene. Then, slides were mounted.

The high resolution images of the stained TMA slides were scanned and captured by

ScanScope scanners. Individual stained TMA spots were visualized and scored for

FOXM1, NBS1 and KIF20A expression using Aperio ScanScope ® system (Aperio technology, USA) by two independent individuals and the average scores were used for calculations.

Product NO. Antibody Species Company Dilution number 1 FOXM1 NBP1-30961 Rabbit Novus biologicals 1:50

2 NBS1 2662 Mouse Cell Signaling 1:50

3 KIF20A Ab104118 Rabbit Abcam 1:50 Table 2.6: List of primary antibodies used for immnohistochemistry

2.15.3 Staining scoring and Statistical Analysis

Tissue sections were scored according to their intensity and percentage of stained tumour cells. As FOXM1, NBS1, and KIF20A were detected in both the cytoplasm and the nucleus, separate evaluation on cytoplasm and nucleus was carried out. The intensity of staining was scored as follows: 1 = weak, 2 = moderate, or 3 = strong.

The percentage of cells positively stained was scored as follows: 1 ≤ 25%, 2 ≤ 50%,

3 ≤ 75%, or 4 > 75%. For each case, a final score was obtained by multiplying the score of intensity with the score of percentage. The total score is the sum of the cytoplasm and nucleus scores.

Statistical analysis was performed using the SPSS programme. The correlation between FOXM1/NBS1, or FOXM1/KIF20A expression were obtained by bi-variate

Pearson Correlation analysis. Survival analysis was estimated by Kaplan Meier analysis with log-rank test. Correlation of FOXM1 expression and clinical-

101 pathological parameters and patient survival were also estimated by multivariate analysis with Cox-regression model.

2.16 Statistic Analysis

The statistical significance of differences between the means of two groups was evaluated by two-tailed unpaired Student’s t-test using the Microsoft Excel programme and GraphPad Prism version 4.03 (GraphPad Software Inc, San Diego,

CA). In all cases, difference were considered to be significant when * P ≤ 0.05, ** P ≤

0.01 and *** P ≤ 0.005 and n.s. for non-significant.

102

CHAPTER 3

THE ROLE OF FOXM1 IN DNA DAMAGE RESPONSE

AND EPIRUBICIN RESISTANCE

103

3.1 Introduction

Anthracyclines (also known as anthracycline antibiotics), including doxorubicin and epirubicin are some of the most effective anticancer agents. Particularly, epirubicin

(Epi) represents a commonly used chemotherapeutic agent in the standard treatment of a wide range of tumours including breast, lung, gastric, and ovary carcinomas, and leukaemia (Minotti et al., 2004). The precise mechanism responsible for the anti-proliferative and cytotoxic effects of epirubicin is not completely understood, but is likely to be involved in inducing DNA intercalation and

DNA double-strand breaks (DSBs), which render cells to undergo apoptosis, mitotic catastrophe, terminal growth arrest, and increased clonogenic tumour cells killed.

Epirubicin effectively blocks the growth and spread of cancer cells and is also associated with a significant reduction in breast cancer mortality; however, intrinsic and acquired chemoresistance is becoming a major obstacle for successful breast cancer management. Elucidation of the molecular mechanisms underlying epirubicin resistance is therefore needed to improve patient survival. As epirubicin exerts its cytotoxic effects by causing DNA damage, the enhancement in DNA repair could be a possible mechanism that enables cancer cells to survive and develop resistance to this DNA damaging agent.

The Forkhead box M1 (FOXM1) transcription factor is responsible for the regulation of a wide range of biological processes, including the cell cycle, cell proliferation, apoptosis, and tumorigenesis (Lam et al., 2013, Myatt and Lam, 2007). Deregulated

FOXM1 has been proposed to be involved in cisplatin and epirubicin resistance, and this could occur through an enhancement of DNA damage repair (Millour et al.,

2011, Kwok et al., 2010a, Monteiro et al., 2013). In response to DNA damage, CHK2 has been report to activate and stabilise FOXM1, which in turn stimulates the

104 expression of DNA repair genes, BRCA2 and XRCC1 (Tan et al., 2007). Recently, results from our laboratory have identified BRIP1 as a new target of FOXM1 in response to DNA damage repair (Monteiro et al., 2013). However, the mechanistic aspects of how FOXM1 modulates DNA repair still remain unclear. Therefore, the aim of this chapter was to investigate the new roles of FOXM1 and its downstream targets in DNA damage response and cellular senescence, which are vital for both the tumour progression and the development of chemotherapeutic drug resistance.

105

3.2 Results

3.2.1 FOXM1-deficient MEFs exhibit increased levels of DNA damage upon epirubicin treatment

Deregulated FOXM1 has been proposed to be involved in cisplatin and epirubicin resistance and this may be due to the enhancement of DNA repair (Millour et al.,

2011, Kwok et al., 2010). However, the precise molecular mechanisms by which

FOXM1 mediates DNA repair events are still not completely understood. To investigate the involvement of FOXM1 in DNA repair mechanism, Foxm1-/- and wild- type (WT) MEFs were either non-treated or treated with 0.1 µM epirubicin, a concentration which does not induce considerable cell death over a short time course (up to 48 h), but instead triggers differential senescence effects on Foxm1-/- and WT MEFs (Monteiro et al., 2013, Khongkow et al., 2014), for 0, 0.5, 2, 4, 24, 48 and 24 h and immunostained for phosphorylated histone H2AX (γH2AX) foci, a well- known marker of DNA double-strand breaks induced by ionising radiation or chemotherapeutic agents (Rogakou et al., 1998, Clingen et al., 2008, Ivashkevich et al., 2012). Analysis of γH2AX foci formation by confocal microscopy revealed that

Foxm1-/- MEFs displayed higher levels of spontaneous DNA breaks upon epirubicin treatment compared to WT MEFs (Figure 3.1). In WT MEFs, maximal foci numbers were reached between 4-24 h of epirubicin treatment. After 24 h of the treatment, as repair progressed there was a significant reduction in the number of γH2AX foci and only a small fraction of the foci persisted for 72 h, indicating that the damaged DNA had been repaired. By contrast, the number of γ-H2AX foci remained considerably higher in Foxm1-/- MEFs, indicating that the depletion of FOXM1 led to a significant increase in DNA damage induced by epirubicin and this could be associated with the

106 impairment of DNA damage repair. Taken together, these results suggest that

FOXM1 may be involved in the repair of DSBs induced by epirubicin.

3.2.2 FOXM1 reconstitution in Foxm1-/- MEFs exhibits the decreased accumulation of γH2AX foci in response to epirubicin treatment

The functional roles of FOXM1 are believe to be mediated by two transcriptionally active isoforms, FOXM1b and FOXM1c (Wierstra and Alves, 2007). However, extensive studies have shown that FOXM1B is the predominant isoform that is overexpressed and highly associated with cancer development, progression and poor prognosis in various cancer cell lines (Lam et al., 2013, Tan et al., 2007, Ye et al., 1997, Costa et al., 2003, Kalinichenko et al., 2004, Myatt and Lam, 2007, Liu et al., 2006, Dai et al., 2007, Li et al., 2009, Xue et al., 2014, Wang et al., 2014,

Nakamura et al., 2010, Gemenetzidis et al., 2009). Therefore, to further confirm the role of FOXM1 in response to DSBs, Foxm1-/- MEFs were transiently transfected with either mCherry-FOXM1B or empty-mCherry control plasmids for 48 h. Then, the transfected cells were either non-treated or treated with 0.1 µM epirubicin for 0, 0.5,

4, 24, 48 and 72 h and immunostained for phosphorylated histone H2AX foci to determine DNA double-strand breaks. The confocal microscopic analysis confirmed that Foxm1-/- MEFs were effectively transfected with mCherry-FOXM1B or empty- mCherry control plasmids (Red). mCherry-FOXM1 protein localised in the nucleus

(Figure 3.2A), whereas mCherry-empty control protein was observed in both the nucleus and cytoplasm of Foxm1-/- MEFs (Figure 3.2B). The analysis of γH2AX foci formation revealed that Foxm1-/- MEFs expressing mCherry-FOXM1B exhibited decreased DNA breaks after epirubicin treatment compared with the neighbouring mCherry-FOXM1 negative cells (Figure 3.2A). However, Foxm1-/- MEFs transfected with the empty-mCherry control have similar kinetics for γH2AX foci accumulation as

107 the mCherry negative cells (Figure 3.2B). Collectively, these results clearly suggest a critical role of FOXM1 in DNA double strand break repair.

Figure 3.1: Foxm1-/- MEFs exhibited higher level of DNA breaks than WT MEFs after epirubicin treatment. WT and Foxm1-/- MEFs cultured on chamber slides were either non-treated (0 h) or treated with 0.1 µM epirubicin for 0.5, 2, 4, 24, 48 and 72 h. Cells were then fixed and immunostained for γH2AX foci (green). Nuclei were counterstained with DAPI (blue). Images were acquired with Leica TCS SP5 (63X magnification). (b) For each time point, images of at least 100 cells were captured and used for quantification of γH2AX foci number. The graph represents average of three independent experiments ± S.D. Statistical analyses were conducted using Student’s t-tests against the correspondent time point (***P≤0.005, significant; ns, non-significant).

108

109

Figure 3.2: Foxm1-/- MEFs expressing ectopic FOXM1B-mcherry plasmids exhibited decreased DNA breaks after epirubicin treatment. Foxm1-/- MEFs were transfected with either ectopic FOXM1- mCherry plasmids. −ve mCherry-FOXM1 negative cells; +ve, mCherry-FOXM1 positive cells (Red) (A) or with mCherry empty vector control plasmids. −ve mCherry negative cells; +ve, mCherry positive cells (Red) (B). 2 days after transfection, cell were then treated with epirubicin for 0, 0.5, 4, 24, 48 and 72 h and immunostained for phosphorylated histone H2AX foci (green) to determine DNA double- strand breaks. Nuclei were counterstained with DAPI (blue). Images were acquired with Leica TCS SP5 (63X magnification). For each time point, images of at least 100 cells were captured and used for quantification of γH2AX foci number. The graph represents average of three independent experiments ± S.D. Statistical analyses were conducted using Student’s t-tests against the correspondent time point (*P≤0.05, ***P≤0.005, significant; ns, non-significant).

110

3.2.3 FOXM1 deletion significantly inhibits long-term clonogenic ability and induces premature senescence in response to DNA damage

From the previous results, I found that Foxm1-/- MEFs accumulated significantly higher numbers of γH2AX foci compared to WT MEFs at prolonged times of 24, 48 and 72 h following epirubicin treatment. Interestingly, accumulation of persistent

γH2AX foci has been considered as one of the important markers for unrepaired or irreparable damaged DNA in senescent cells. These suggested that FOXM1 may protect cells from genotoxic agent-induced cell death or senescence by enhancing

DNA damage repair mechanisms. However, the role of FOXM1 in promoting long- term clonogenic survival in response to chronic DNA damage induced by epirubicin treatment has not yet been investigated. To address this, WT and Foxm1-/- MEFs were treated with a wide range of epirubicin doses (0-100 nM) and their long-term viability was investigated by clonogenic assay (Figure S3). The results showed that at 0, 20 and 40 nM epirubicin, the colony formation capacity of Foxm1-/- MEFs was significantly impaired compared with WT MEFs (Figure 3.3A, Figure S3).

Interestingly, previous studies from our laboratory have shown that these low epirubicin concentrations (20 or 40 nM) do not induce considerable cell death over shorter time courses (up to 48 h) (Millour et al., 2011, Monteiro et al., 2013). I therefore hypothesised that low levels of DNA damage induced by epirubicin may trigger senescence-associated anti-proliferative responses instead of activating apoptosis. To test this hypothesis, WT and Foxm1-/- MEFs were assayed for senescence-associated β-galactosidase (SA-βgal) activity, a well-established biomarker of senescence, following low doses of epirubicin treatment (Figure 3.3B).

At 20 and 40 nM epirubicin treatment, significantly higher numbers of Foxm1-/- MEFs

111 displayed distinct SA-βgal activity and ‘flat cell’ morphology compared with WT

MEFs.

To confirm this further, WT and Foxm1-/- MEFs were exposed to moderate levels of ionising-radiation (Figure 3.3C-D). At 5Gy γ-irradiation, the colony formation capacity of Foxm1-/- MEFs was also significantly reduced compared to WT MEFs (Figure

3.3C). Consistently, Foxm1-/- MEFs also displayed higher percentages of senescent cells compared with WT MEFs following exposure to 5 Gy of γ-irradiation (Figure

3.3D), suggesting that FOXM1 has a key role in protecting the cell against DNA damage-induced senescence.

112

Figure 3.3: FOXM1 deletion inhibits clonogenic growth and induces cellular senescence in response to DNA damage in MEFs. (A) Clonogenic assays were performed to assess the colony formation efficiency of Foxm1-/- and WT MEFs. 2,000 cells were seeded in six well plates, treated with 0, 20 and 40 nM of epirubicin and grown for 15 days. The cells were then stained with crystal violet (left panel). The result (right panel) represents average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01). (C) Clonogenic assay of Foxm1-/-and WT MEFs were either non-irradiated or exposed to 5 Gy of γ- irradiation. (B) SA-βgal staining of Foxm1-/-and WT MEFs treated with 0, 20 and 40 nM of epirubicin. 5 days after treatment, cells were stained for SA-β-galactosidase activities. (D) SA-βgal staining of Foxm1-/- and WT MEFs treated with irradiation (0 and 5 Gy). The graphs (B, right panel) and (D, right panel) show the percentage of SA-βgal positive cells as measured from five different fields from two independent experiments. Bars represent average ± SD. Statistical significance was determined by Student’s t-test (**P ≤ 0.01, significant; ns, non-significant).

113

3.2.4 Knockdown of FOXM1 leads to increased levels of DNA damage in MCF-7

EpiR cells.

In order to investigate whether FOXM1 confers epirubicin resistance through the enhancement of DNA damage repair, I decided to assess the effect of silencing

FOXM1 using siRNA targeting all isoforms of FOXM1 in the previously established epirubicin-resistance MCF-7 (MCF-7 EpiR) breast cancer cells, which were generated by the repeated passages of MCF7 cells in increasing concentrations of epirubicin

(Millour et al., 2011). MCF-7 EpiR cells cultured on slide chambers were transiently transfected with either NS or FOXM1 siRNA for 24 h. The transfected cells were then treated with 1 µM epirubicin, a clinical concentration generally used in cancer therapy (Iwakiri et al., 2008, Millour et al., 2011), and fixed at the indicated time points (0, 4, 24, 48, and 72 h) prior to immunostaining for γH2AX. The γH2AX foci formation analysis revealed that knockdown of endogenous FOXM1 expression in

MCF-7 EpiR cells displayed a significant increase in DNA breaks after epirubicin treatment (Figure 3.4). The induction of γH2AX foci was observed following 4 h of epirubicin treatment and remained persistently high through the entire time points, suggesting that diminished FOXM1 expression caused a deficiency in DNA repair which leads to an increase in DNA breaks. This result indicates that depletion of

FOXM1 is sufficient to render epirubicin resistant MCF-7 EpiR cells more susceptible to epirubicin-induced DSBs.

114

Figure 3.4: Knockdown of FOXM1 leads to increased levels of DNA damage in MCF-7 EpiR cells. (A) MCF-7 EpiR cells were transfected with control siRNA (NS siRNA) or with FOXM1 siRNA for 24 h. Cells were cultured on chamber slides and treated with 1 μM of epirubicin for 0, 4, 24, 48 and 72 h and then stained for γH2AX (green) and DAPI (blue). (B) The graph below shows quantification of γH2AX foci number. Bars represent an average of three independent experiments ± S.D. Statistical analyses were conducted using two-tailed unpaired Student’s t-test against the correspondent time point (***P≤0.0001, significant; n.s., non-significant). (C) The silencing effect of FOXM1 siRNA was detected by western blot analysis.

115

3.2.5 Knockdown of FOXM1 significantly sensitises MCF-7 breast cancer cells to epirubicin-induced cellular senescence

The effects of FOXM1 depletion by siRNA on the long-term viability of MCF-7 and the MCF-7 EpiR breast cancer cell lines were next examined by clonogenic assays

(Figure 3.5). FOXM1-knockdown sensitised MCF-7 cells to long-term proliferative arrest at relatively low epirubicin concentrations (10 and 20 nM) (Figure 3.5A), whereas at higher concentrations the colony formation ability was completely lost in

MCF-7 cells. Notably, FOXM1-depletion alone almost completely impaired the colony formation ability of MCF-7 EpiR cells irrespective of the dosage of epirubicin used (Figure 3.5C), suggesting that MCF-7 EpiR cells are dependent on high FOXM1 expression for long-term survival. Similar to FOXM1 deletion in MEFs, FOXM1 knockdown in MCF-7 cells also enhanced the number of cells exhibiting SA-βgal activity and changed in morphology at 0 and 10 nM epirubicin (Figure 3.5B).

Consistently, FOXM1 knockdown in MCF-7 EpiR cells resulted in almost all cells displaying senescence-associated SA-βgal activity and morphology independent of epirubicin levels (Figure 3.5D), suggesting that MCF-7 EpiR cells have become dependent on FOXM1 to override senescence. As mentioned earlier, senescent cells accumulate γH2AX foci as markers for unrepaired or irreparable damaged DNA. To confirm this further, I studied the effects of FOXM1 knockdown on nuclear γH2AX foci formation in MCF-7 and MCF-7 EpiR cells following 5 Gy of γ-irradiation, a radiation dosage which induced senescence in MCF-7 cells (Karimi-Busheri et al.,

2010). Quantification of γH2AX foci showed that γ-irradiation induced significantly higher number of foci formation over the prolonged time course of 24, 48 and 72 h in

FOXM1 depleted cells in comparison to control siRNA-transfected MCF-7 (Figure

116

3.6A) and MCF-7 EpiR cells (Figure 3.6B), further indicating that FOXM1 has a critical role in overcoming DNA damage-induced senescence.

Figure 3.5: Knockdown of FOXM1 suppresses cell growth and induces cellular senescence in MCF-7 and MCF-7 EpiR cells. A) MCF-7 and (C) MCF-7 EpiR were transfected with NS siRNA or FOXM1 siRNA. Twenty-four hours after transfection, 2,000 cells were seeded in six well plates, treated with epirubicin, grown for 15 days and then stained with crystal violet (left panel). The graphs (A; right panel) and (C; right panel) represent average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, significant;

117 ns, non-significant). In parallel, (B) MCF-7 and (C) MCF-7 EpiR transfected with NS siRNA or siFOXM1 were seeded in six well plates, treated with epirubicin. 5 days after treatment, cells were stained for SA-β-galactosidase activity. The graphs (B; right panel) and (D; right panel) show the percentage of SA-β-galactpsidase-positive cells as measured from five different fields from two independent experiments. Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, significant; ns, non-significant).

Figure 3.6: Knockdown of FOXM1 causes the accumulation of γH2AX foci in response to γ- irradiation. (a) MCF-7 and (b) MCF-7 EpiR were transfected with NS siRNA or FOXM1 siRNA. Twenty-four hours after transfection, cells were either non-exposed or exposed to 5 Gy of γ-irradiation for 24, 48 and 72 h. Cells were then fixed and immunostained for γH2AX foci (green). Nuclei were counterstained with 4′-6-diamidino-2-phenylindole (DAPI; blue). Images were acquired with Leica TCS SP5 (63X magnifications). For each time point, images of at least 100 cells were captured and used for quantification of γH2AX foci number. Results represent average of three independent experiments±S.D. Statistical analyses were conducted using two-tailed unpaired Student’s t-test against the correspondent time point (**P ≤ 0.01, significant; n.s., non-significant).

118

3.2.6 FOXM1 and DNA repair gene, NBS1, are up-regulated in MCF-7 EpiR cells

Previous studies in our laboratory have shown that FOXM1 is essential for homologous recombination (HR) DNA repair activity after DSBs (Monteiro et al.,

2013). In order to elucidate in detail the mechanistic role of FOXM1 in DNA damage- induced senescence and epirubicin resistance, I next compared the expression patterns of FOXM1 and the main HR repair proteins, including NBS1, MRE11,

RAD50 and ATM (Bartkova et al., 2008) in response to epirubicin treatment in both parental MCF-7 cells and MCF-7 EpiR cells. Western blot analysis showed that in

MCF-7 cells there was a transient induction of FOXM1 protein at 4 h post epirubicin treatment before decreasing gradually over the treatment time. By contrast, the

MCF-7 EpiR cells expressed a significantly higher level of FOXM1 compared to the parental MCF-7 cells and this high level of FOXM1 was maintained throughout the timecourse treatment (Figure 3.7A, Figure S4). Furthermore, the expression levels of

DNA repair proteins; NBS1, MRE11, RAD50, and ATM were observed at higher levels in the MCF-7 EpiR cells compared with the expression levels in MCF-7 cells.

Interestingly, I also found that NBS1 as well as its phosphorylated form, P-NBS1, exhibited similar kinetics to FOXM1 in response to epirubicin in both MCF-7 and

MCF-7EpiR cells (Figure 3.7A, Figure S4). Consistent with these observations, RT- qPCR analysis also revealed that NBS1 mRNA levels increased 2 fold in MCF-7

EpiR cells compared to the parental MCF-7 cells which exhibited decreased NBS1 transcriptional level by almost 50% after 24 h epirubicin treatment (Figure 3.7B).

These results suggest that FOXM1 may transcriptionally regulate NBS1 expression and this regulation could mediate epirubicin resistance in breast cancer cells.

119

Figure 3.7: Epirubicin resistant cell line exhibits increased FOXM1 and NBS1 protein and mRNA expression levels (A) Representative western blot of three independent experiments, determining the protein expression levels of FOXM1, NBS1, P-NBS1, MRE11, RAD50, P-ATM, and ATM in MCF-7 and MCF-7 EpiR cell lines after being treated with 1 µM of epirubicin at different time points. β-tubulin was used as a loading control. (B) NBS1 mRNA transcript level was carried out by RT-qPCR analysis and normalised to the ribosomal protein L19 RNA expression. Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, ***P≤0.005, significant; ns, non-significant).

120

3.2.7 FOXM1 enhances NBS1 expression and ATM activity to promote DNA damage repair signalling

To test whether NBS1 is a downstream target of FOXM1 required for conferring epirubicin resistance in breast cancer cells, both MCF-7 and MCF-7 EpiR cells were depleted of FOXM1 by employing FOXM1 siRNA transfection for 24 h.

Subsequently, cells were treated with 1 µM epirubicin for 24 h and harvested for western blot and RT-qPCR analysis. Western blot and RT-qPCR analysis showed that FOXM1 knockdown caused a significant decrease in the NBS1 protein and mRNA levels in both MCF-7 and MCF-7 EpiR cells, whereas only a significant reduction was observed for the RAD50 and MRE11 proteins (Figure 3.8A, Figure

S5), but not at the mRNA level, upon FOXM1 knockdown (Figure 3B). This suggests that in contrast to NBS1, RAD50 and MRE11 are not direct downstream targets of

FOXM1, raising the possibility that their protein expression levels might be stabilised through the formation of an active MRN complex. NBS1 have been recently shown to function upstream of ATM activation in addition to its well-established role downstream of ATM (Uziel et al., 2003, Wu et al., 2012, Horejsi et al., 2004, Lee and

Paull, 2007, You et al., 2005). From the western blot analysis, it is also notable that

P-ATM levels were strongly induced by epirubicin treatment, but this induction was abolished after FOXM1 silencing by siRNA in both MCF-7 and MCF-7 EpiR cells

(Figure 3.8A, Figure S5), suggesting that FOXM1 may be involved in the regulation of NBS1 expression and ATM activity.

121

Figure 3.8: MCF-7 and MCF-7 EpiR cells depleted in FOXM1 expression revealed significant decreases in both protein and mRNA levels of NBS1. MCF-7 and MCF-7 EpiR cells were either transfected with non-specific control siRNA or siRNA against FOXM1 for 24 h. After transfection, cells were treated with 1 µM epirubicin for 0 and 24 h. (A) Representative western blot of three independent experiments, determining the protein expression levels of FOXM1, NBS1, P-NBS1, MRE11, RAD50, P-ATM, and ATM. β-tubulin was used as a loading control. (B) RT-qPCR was performed to indicate the relative FOXM1, NBS1, MRE11 and RAD50 mRNA transcriptional level. Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, ***P≤0.005, significant; ns, non-significant).

122

3.2.8 FOXM1-deficient cells display decreased expression of the DNA repair gene NBS1

To confirm further that FOXM1 modulates ATM activity through regulating NBS1 expression, WT and Foxm1-/- MEFs were treated with 0.1 µM epirubicin for 0, 0.5, 2,

4, 16, and 24 h. The expression levels of FOXM1, P-ATM, P-NBS1 and NBS1 were subsequently investigated by western blot analysis (Figure 3.9). Consistent with the results from the breast cancer cell lines, NBS1 and P-NBS1 were modulated with similar kinetics as FOXM1 in WT MEFs following epirubicin treatment, but were barely detectable in Foxm1-/- MEFs. By contrast, RAD50 and MRE11 were expressed at only marginally lower levels in Foxm1-/- MEFs compared to WT MEFs.

The western blot results also showed that although ATM was expressed at comparable levels in WT and Foxm1-/- MEFs, P-ATM was induced by epirubicin and expressed at much higher levels in WT compared to Foxm1-/- MEFs (Figure 3.9A).

Consistently, analysis of RT-qPCR data also indicated that upon epirubicin treatment the NBS1 and FOXM1 mRNA displayed similar kinetics in WT and Foxm1-/- MEFs, further confirming FOXM1 regulates NBS1 expression (Figure 3.9B). Together these findings led us to postulate that FOXM1 may target NBS1 expression at the transcriptional level to enhance DNA damage repair signalling and epirubicin resistance.

123

Figure 3.9: FoxM1-deficient MEFs exhibited decreased mRNA and protein levels of the DNA repair gene NBS1 compared with wild type MEFs. Wild type and Foxm1-/- MEFs were treated with 0.1 µM epirubicin for 0, 1, 2, 4, 8, 24, and 48 h and the expressions of FOXM1 and NBS1 were determined by western blot and RT-qPCR analysis.

124

3.2.9 Overexpression of FOXM1 leads to upregulation of NBS1 expression and enhances ATM activity.

I next examined the effect of ectopic expression of FOXM1 on the NBS1 expression and ATM activity. To this end, MCF-7 cells were transiently transfected with either an empty vector or FOXM1B-pcDNA3 plasmids. Twenty-four hours after transfection, cells were treated with 1 µM epirubicin for different time periods (0, 4, 24, and 48 h).

The expression levels of NBS1 protein and ATM activity were then evaluated by western blot analysis. The results showed that overexpression of FOXM1 enhanced the expression of NBS1 protein and ATM phosphorylation in MCF-7 cells (Figure

3.10), further confirming that FOXM1 regulates NBS1 expression and thus MRN complex formation to promote ATM activation and phosphorylation. It is notable that despite FOXM1 overexpression, NBS1 levels decreased after 24 h of epirubicin treatment. This is probably because both FOXM1 and NBS1 are also regulated at post-transcriptional levels in response to epirubicin (de Olano, 2012).

Figure 3.10: FOXM1 regulates NBS1 expression and modulates ATM phosphorylation. MCF-7 cells were transfected with pcDNA3 empty vector or FOXM1-pcDNA3 plasmids following with these cells were subjected to 1 µM epirubicin treatment for 0, 4, 24 and 48 h. (A) Western blots were performed to analyse the protein expression level changes of FOXM1, NBS1, P-ATM, ATM, RAD50, MRE11 and β-Tubulin was used as a loading control. (B) Right panel show quantification of proteins density using ImageJ software. Each sample was normalised to β-Tubulin (n=2).

125

3.2.10 FOXM1 regulates NBS1 expression through a FHRE in its promoter

To investigate if the regulation of NBS1 by FOXM1 also occurs at the promoter level,

MCF-7 cells were transiently co-transfected with increasing amounts of the FOXM1 expression construct and the luciferase reporter gene under the control of either a human NBS1 wild-type (WT) or a mutant (mut) NBS1 (1.5 Kbp) promoter with a putative FHRE (forkhead response element) (-78 bp) mutated (Figure 3.11). The results showed that the (WT) NBS1 promoter activity was augmented by FOXM1 in a dose-dependent manner, whereas the mutant (mut) NBS1 promoter had lower basal promoter activity and failed to be induced by FOXM1. Collectively, these results suggest that FOXM1 is able to transactivate NBS1 gene through the FHRE located at position -78 bp, providing evidence that NBS1 is a direct target gene of FOXM1.

Figure 3.11: FOXM1 regulates NBS1 expression through a FHRE site within its promoter. MCF-7 cells were transiently co-transfected with 20 ng of NBS1-WT or NBS1-mutant promoter plasmid (pGL3 -Basic Vector) together with increasing concentrations (0, 10, 15 and 20 ng) of pcDNA3-FOXM1 plasmid and transfection control Renilla plasmid (pRL-TK). After 24 h, cells were assayed for Firefly and Renilla luciferase activity. The relative luciferase activity was calculated after normalising with Renilla activity.

126

3.2.11 FOXM1 and NBS1 required for efficient HR-mediated repair

Recent studies from our laboratory have shown that FOXM1 is essential for homologous recombination (HR) but not non-homologous end joining (NHEJ) DNA repair pathway in response to DSBs (Monteiro et al., 2013). To determine the potential role of FOXM1 and NBS1 in DSB repair via HR, HeLa cells containing a

DR-GFP reporter construct (Morris et al., 2009, Pierce et al., 1999) were transfected with either FOXM1 siRNA, NBS1 siRNA, pFOXM1 or pNBS1 expression plasmids.

Subsequently, these cells were transfected with I-Scel expression plasmid. Transient expression of I-SceI endonuclease generates a DSB at the integrated GFP gene sequences and stimulates HR repair. Three days after I-Scel transfection, GFP positive cells were monitored by flow cytometry. For each experiment, 50,000 cells were analysed per treatment group and the frequency of HR events was calculated from the number of GFP-positive cells divided by the total number of cells analysed

(Figure 3.12A). Western blot analysis was used to confirm the transfection efficiency

(Figure 3.12B). The results showed that ectopic expression of either FOXM1 or

NBS1 significantly enhanced the DNA repair via HR by 82.9% and 87.3%, respectively (Figure 3.12C and 12D). In contrast, the depletion of FOXM1 or NBS1 using siRNA led to a significant reduction in HR frequency when compared with the cells transfected with non-specific control siRNA (54.0% and 67.6%, respectively)

(Figure 3.12C and 12D). These data suggest that both FOXM1 and NBS1 play an important role in HR repair. Next, to explore the functional link between FOXM1 and

NBS1 in HR-mediated DSB repair, HeLa cells carrying DR-GFP were co-transfected with FOXM1 plasmid and NBS1 siRNA. The results revealed that overexpression of

FOXM1 and depletion of its proposed downstream targets, NBS1, caused a significant reduction in HR repair efficiency (49.4%) (Figure 3.12C and 12D).

127

Surprisingly, this reduction was not significantly higher than that observed in NBS1 depletion alone, indicating that the function of FOXM1 to direct DSBs towards HR is related with NBS1 expression.

Figure 3.12: FOXM1 and NBS1 are required for efficient HR repair (A) Schematic of DR-GFP assay for DSB repair by homologous recombination. The SceGFP reporter is a GFP gene which contains an I-SceI endonuclease site within the coding region. Cleavage of the I-Scel site by exogenously expressed I-Scel and repair induced by HR leads to GFP expressing cells. (B) Transfection efficiency in HeLa cells were monitored by western blot analysis. (C) The percentage of GFP positive cells was determined by flow cytometry. (D) Comparison of homologous recombination repair in HeLa cells carrying DR-GFP. Bars represent average ± SD of three independent experiments. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P ≤ 0.05, significant; ns, non-significant).

128

3.2.12 Ectopic expression of NBS1 in Foxm1-/- MEFs reduces the sensitivity to epirubicin, as indicated by the decreased accumulation of γH2AX foci

To further demonstrate that downregulation of NBS1 expression in FOXM1-depleted cells is responsible for the accumulation of senescence associated γH2AX foci in response to epirubicin treatment. I next determined whether the overexpression of

NBS1 in Foxm1-/- MEFs is able to reduce the accumulation of γH2AX foci particularly at the longer times of 24, 48 and 72 h. To this end, Foxm1-/- MEFs were transiently transfected with empty-mCherry alone (control) or co-transfected mCherry with

NBS1 expression plasmid for 48 h. Then, the transfected cells were either non- treated or treated with 0.1 µM epirubicin for 0, 0.5, 4, 24, 48 and 72 h and immunostained for γH2AX foci to determine DNA double-strand breaks. The analysis of γH2AX foci formation revealed that Foxm1-/- MEFs expressing NBS1 displayed a significantly lower number of γH2AX foci compared with the neighboring untransfected cells at the longer time points (24, 48 and 72 h) after epirubicin treatment (Figure 3.13). Interestingly, the overexpression of NBS1 had the same effects as reconstituting Foxm1-/- MEFs with mCherry-FOXM1 (Figure 3.2A). By contrast, Foxm1-/- MEFs transfected with the empty-mCherry control have similar kinetics for senescence-associated γH2AX foci accumulation as the non-transfected cells (Figure 3.13 and Figure 3.2). Taken together, these findings clearly confirm that the role of FOXM1 in HR-mediated DSB repair is related to its ability to control NBS1 expression.

129

Figure 3.13: Ectopic expression of NBS1 in Foxm1-/- MEFs reduces the accumulation of γH2AX foci. Foxm1-/- MEFs were either transfected with mCherry control plasmids or co-transfected with mCherry and NBS1 plasmids (Red) and treated with 0.1 µM of epirubicin for 0, 0.5, 2, 4, 24, 48 and 72 h. The cells were then immunostained for γH2AX foci (Green) and nuclei were counterstained with DAPI (blue) to determine DNA double-strand breaks. γH2AX foci quantification is shown. Bars represent the average of γH2AX foci per cell from three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P≤0.05, ***P≤0.005, significant; ns, non-significant).

130

3.2.13 Depletion of NBS1 increases sensitivity to epirubicin and induces senescence associated phenotypes in MCF-7 breast cancer cells

To investigate the potential role of NBS1 in epirubicin resistance, I depleted NBS1 expression using siRNA in both MCF-7 and MCF-7 EpiR cells and studied its effects on cell proliferation using the SRB assay. Interestingly, the result showed that knockdown of NBS1 revealed dose-dependent cell growth inhibition in both breast cancer cell lines, MCF-7 and MCF-7 EpiR, induced by epirubicin (Figure 3.14). As

FOXM1 is involved in DNA damage-induced senescence after long-term treatments with low concentrations (10 or 20 nM) of epirubicin and also controls NBS1 expression, the effects of NBS1 depletion on the long-term viability of MCF-7 and

MCF-7 EpiR cells were then examined by clonogenic assays. Similar to FOXM1 depletion, NBS1-knockdown sensitised MCF-7 and MCF-7 EpiR cells to long-term proliferative arrest following treatment with epirubicin (Figure 3.16A and 16C, respectively). Consistent with this, NBS1 knockdown in MCF-7 and MCF-7 EpiR cells also enhanced the number of cells exhibiting senescence-associated β-gal activity and flat morphology at 0 and 10 nM epirubicin (Figure 3.16B and 16D). These findings strongly suggest that NBS1 protects MCF-7 and MCF-7 EpiR cells from undergoing cellular senescence in response to DNA damage induced by epirubicin.

Notably, differential effects of NBS1 depletion in the absence of epirubicin treatment on cell proliferation were observed only in long term proliferation assay as indicated in the 0uM epirubicin condition in Figure 3.16.

Next, I studied the effect of NBS1 depletion in the resistant MCF-7 EpiR cells using western blot analysis. The result revealed that depletion of NBS1 led to a decrease in P-ATM and P-NBS1 expression (Figure 3.15), further supporting the idea that

NBS1 is needed for ATM activation in response to epirubicin. Consistent with this

131 finding, the previous study in the lab also showed that reconstitution of NBS1 expression in NBS1-deficient NBS1-LBI cells can restore the activation of ATM in response to epirubicin (Khongkow et al., 2014).

Figure 3.14: The depletion of NBS1 increases cell sensitivity to epirubicin. MCF-7 (A) and MCF- 7 EpiR (B) cells were transfected with either NS siRNA (control) or NBS1 siRNA. 24 h after transfection, cells were treated with increasing concentrations of epirubicin for 48 h. Subsequently, rates of cell proliferation were analysed by SRB assay. Data shown are an average of three independent experiments± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P≤0.05, **P ≤ 0.01, significant; ns, non-significant). Western blot analysis was used to monitor the transfection efficiency of NBS1 siRNA in MCF-7(C) and MCF-7 EpiR (D) cells.

132

Figure 3.15: MCF-7 EpiR cells depleted in NBS1 expression revealed a decrease in P-ATM and P-NBS expression. MCF-7 EpiR cells were either transfected with non-specific control siRNA or siRNA smart pool against NBS1 for 48 h. After transfection, cells were treated with 1 µM epirubicin for 0, 4, 24 and 48 h. The protein expression levels of FOXM1, NBS1, MRE11, and RAD50, ATM, P-ATM and Cleaved PARP were determined by western blot analysis.

133

134

Figure 3.16: NBS1 depletion inhibits cell growth and induces cellular senescence in MCF-7 breast cancer cells. MCF-7 (A) and MCF-7 EpiR (C) were transfected with NS siRNA or NBS1 siRNA. 24 hours after transfection, 2,000 cells were seeded in six well plates, treated with epirubicin, grown for 15 days and then stained with crystal violet (left panel). The result (right panel) represents average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, significant; ns, non-significant). In parallel, MCF-7 (B) and MCF- 7 EpiR (D) transfected with NS siRNA or siNBS1 were seeded in six well plates, treated with epirubicin for 5 days. Cells were stained for SA-β-galactosidase activities. The graphs show the percentage of SA-β-galactpsidase-positive cells as measured from five different fields from two independent experiments. Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test test (**P ≤ 0.01, ***P≤0.005, significant; ns, non-significant).

135

3.2.14 Overexpression of FOXM1 or NBS1 confers resistance to epirubicin possibly by enhancing DNA repair pathway in MCF-7 cells

In Figures 3.5 and 3.14, I have shown previously that NBS1 or FOXM1 depletion is able to render MCF-7 and MCF-7 EpiR cells more sensitive to epirubicin-induced cellular senescence. As a consequence, cells overexpressing FOXM1 or NBS1 should be less sensitive to epirubicin treatment. To test this hypothesis, MCF-7 cells were transiently transfected with control empty vector, pmCherry-FOXM1 or pmCherry-NBS1 for 24 h, and then treated with increasing concentrations of epirubicin for 48 h and their cell viability measured by SRB proliferation assay.

Transfection efficiency was examined by western blot analysis as shown in Figure

3.17A. The result of SRB assay revealed that the overexpression of FOXM1 or

NBS1 alone was sufficient to confer epirubicin resistance in MCF-7 cells (Figure

3.17B). Furthermore, cells transfected with the pmCherry-FOXM1 plasmid (stained in red) appear to show decreased levels of DNA damage induced by epirubicin (γH2AX foci stained in green) compared to non-transfect cells. These data suggest that overexpression of FOXM1 and NBS1 could protect breast cancer cells from epirubicin-induced senescence by enhancing DNA damage repair. However, this experiment was performed only two times, thus there is a need to repeat more experiments in order to obtain sufficient sample size for statistical analysis.

136

110

Figure 3.17: Overexpression of FOXM1 or NBS1 promotes epirubicin resistance and decrease DNA damage in MCF-7 cells. MCF-7 cells were transfected with control empty vector, pmCherry- FOXM1 or pmCherry-NBS1 for 24 h. Overexpression levels of FOXM1 and NBS1 were confirmed by western blot analysis (A). Cells were then treated with increasing concentrations of epirubicin for 48 h and their cell viability measured by SRB assay (B). Representative data from three independent experiments are shown (n=2). (C) MCF-7 cells were transfected with the control empty vector or with pmCherry-FOXm1for 24 h. Cells were then seeded on 4-chamber slides and treated with 1 µM epirubicin at 0 and 24 h, as indicated, and stained for yH2AX in green (AlexaFluor 488) , DAPI in blue and mCherry in red.

137

3.2.15 Correlation between NBS1 and FOXM1 expression in breast cancer samples

After establishing that NBS1 is a direct transcriptional target of FOXM1 in breast cancer cell lines and mouse embryonic fibroblasts, in collaboration with Professor

Khoo Ui Soon (Department of Pathology, the University of Hong Kong), we analysed the correlation of FOXM1 and NBS1 expression on tissue microarray (TMA) slides containing 116 human breast cancer patient samples (Chen et al., 2010) using immunohistochemistry (Figure 3.18A). The results indicated that FOXM1 and NBS1 protein were detected in both nuclear and cytoplasmic compartments. Statistical analysis of the expression patterns revealed that there was a strong highly significant correlation between FOXM1 nuclear staining and total NBS1 staining (Pearson coefficient=0.318, p=0.002), providing further physiological evidence that FOXM1 regulates NBS1 expression in breast cancer patient samples (Figure 3.18A). Survival analysis by Kaplan-Meier estimate with log-rank test revealed that the high nuclear expression of NBS1 was correlated with overall poor survival (p=0.164) (Figure

3.18B). In addition, multivariate survival analysis using Cox's regression model, when adjusted by patients’ T-stage and lymph-node involvement, indicated that

NBS1 expression was significantly correlated with poorer survival (RR=2.869, p=0.048). Further analysis of FOXM1 and NBS1 mRNA transcript expression in another previously published cohort (2878 breast cancer patients) (Györffy et al.,

2010) confirmed that high FOXM1 and NBS1 mRNA expression levels were very significantly associated with poor survival (p<0.0001 and p=0.0012, respectively for overall survival, Kaplan-Meier analysis) (Figure 3.19). The significance of NBS1 in survival analysis suggests a direct involvement of NBS1 in regulating cell senescence and DNA damage repair in genotoxic drug response.

138

Figure 3.18: Correlation between NBS1 and FOXM1 expression in breast cancer samples. A) Correlation of FOXM1 and NBS1 expression were assessed by immunohistochemistry in breast cancer samples (N=116). FOXM1 and NBS1 staining were detected in both nuclear and cytoplasmic compartments. NBS1 high and low are categorized according to its median score, so half of the samples are NBS1 high and half of samples are NBS1 low. Statistical analysis of the expression patterns showed that there was a strong and significant correlation between FOXM1 nuclear staining and total NBS1 staining. B) Kaplan–Meier analysis of overall survival for NBS1 staining with log-rank test indicated that there was a statistically non-significant trend linking high NBS1 expression to poor survival in breast cancer patients.

139

Figure 3.19: Kaplan-Meier analysis of overall survival for FOXM1 and NBS1 mRNA transcript expression in a previously web-based published cohort (Györffy et al., 2010). High FOXM1 and NBS1 mRNA expression levels are very significantly associated with poor survival (p<0.0001 and p=0.0012, respectively).

140

3.3 Discussion

Epirubicin (Epi) represents a widely used chemotherapeutic agent in the clinical treatment of breast cancer (Minotti et al., 2004). The precise mechanism of action of epirubicin is likely to be associated with the generation of DNA double-strand breaks

(DSBs), which lead cells to apoptosis and increased tumour cell death (Choudhury,

Zhao et al. 2009). However, the failure of epirubicin treatment occurs as a result of development of drug resistance and systemic toxicity, because high doses are required to obtain a significant anticancer effect. Interestingly, my work in this chapter shows that breast cancer cell lines as well as mouse embryonic fibroblasts can be induced to enter cellular senescence by DNA damaging agents including epirubicin and ionising radiation. Notably, the concentrations of DNA damaging agents required for inducing senescent phenotypes in these cells are substantially lower than that required to induce cell death, suggesting the possibility that cellular senescence may be the predominant mechanism of action for genotoxic anti-cancer agents (Millour et al., 2011, Monteiro et al., 2013).

The development of epirubicin-acquired resistance represents a major obstacle to successful treatment of breast cancer. One of the crucial factors in the development of chemoresistance is the enhanced DNA repair ability of cancer cells which prevents the cells accumulating DNA damage induced by DNA damaging agents

(Stewart, 2007, Perez, 1998, Fink et al., 1996). Although several recent papers have already addressed the role of FOXM1 in genotoxic drug resistance, its exact role in epirubicin resistance are still not yet completely understood (Koo et al., 2012,

Monteiro et al., 2013, Millour et al., 2011, Kwok et al., 2010b). In this chapter, I found that FOXM1 plays a central role in modulating DNA damage-induced senescence and epirubicin resistance. Depletion of FOXM1 sensitised MCF-7 breast cancer cells

141 as well as MEFs cells into entering epirubicin-induced senescence, as indicated by cellular markers of senescence, including the accumulation of γH2AX foci, the loss in long-term cell proliferation, the induction of β-galactosidase activity as well as the flat cell morphology (Figure 3.1, 3.3, 3.4 and 3.5). These results were consistent with the previous observation that FOXM1-deficient cells exhibited high level of γH2AX foci, enhanced p53 transcription activity and increased expression of the p21 gene in response to γ-irradiation (Tan et al., 2005). However, in Figure 3.1, there was no significant difference in the level of H2AX foci between WT and Foxm1-/- MEFs at basal level (untreated condition). This may be due to the immortalization of Foxm1-/-

MEFs. Conversely, reintroduction of FOXM1 in FOXM1-deficient MEFs decreased the accumulation of senescence-associated γH2AX foci (Figure 3.2). Crucially, I have also shown a novel role of FOXM1 in mediating DNA-damage induced senescence and epirubicin resistance through the direct regulation of NBS1 expression and ATM activity. NBS1 is a part of the MRE11-RAD50-NBS1 (MRN) complex, which is the main element in cellular response to DSBs including damage sensing and activation of cell cycle checkpoints signalling pathways (Kang et al.,

2005, Kobayashi et al., 2004, Saito et al., 2013). Knockdown of FOXM1 in MCF-7 and MCF-7 EpiR cell lines led to a significant downregulation of NBS1 at both protein and mRNA levels, whereas overexpression of FOXM1 enhanced NBS1 expression and ATM activity in response to epirubicin treatment (Figure 3.8). These results support the notion that FOXM1 is a crucial transcriptional regulator of the DNA repair gene NBS1 in response to DNA damage. Moreover, I found that FOXM1 is upregulated in MCF-7 EpiR cells compared with the parental sensitive MCF-7 cells.

FOXM1 expression is downregulated in response to epirubicin in MCF-7 cells, but remains persistently high in the resistant cells following epirubicin treatment. These

142 data suggest that FOXM1 is a target of epirubicin and that it has a role in mediating epirubicin action and resistance. Similar to FOXM1, NBS1 was also overexpressed in epirubicin resistant MCF-7 cells (Figure 3.7). However, there were slight discrepancies between the protein expression of FOXM1 in response to epirubicin treatment in Figure 3.7 and 3.8. This might be due to the effect of cell-cycle, we didn’t synchronize the cells, therefore; most cells are at the different stage of cell cycle and this might cause the different cellular outcome in response to epirubin treatment.

Interestingly, the depletion of FOXM1 in the resistant cancer cells not only resulted in the downregulation of the protein levels of NBS1 but also RAD50 and MRE11 indicating that the MRN complex might be regulated by FOXM1. However, these trends were not reflected at the mRNA level (Figure 3.8A and 8B). The results suggest the possible post-transcriptional regulation of MRN complex by FOXM1 in response to DNA damage repair as well as drug resistance in breast cancer cells.

NBS1 appears to act both upstream and downstream of ATM in the DNA damage response pathway (You et al., 2005, Uziel et al., 2003, Horejsi et al., 2). The interactions of NBS1 with ATM is critical for ATM activation to be functional in response to DNA damage (Cariveau et al., 2007 and Kang et al., 2005, Saito et al.,

2013). Another interesting observation drawn from this chapter is that the increased number of DNA breaks in Foxm1−/− MEFs and FOXM1-depleted MCF-7 cells after epirubicin treatment correlated with decreased activation of ATM, which is the main player in the DSB repair pathway. These events resulted in impaired cellular DNA- damage response and enhanced the sensitivity to DNA-damaging agents determined by an apoptosis marker, cleaved-PARP (Figure 3.15). This further supports that FOXM1 directly regulates NBS1 expression and, in turn, upregulation

143 of NBS1 indirectly enhances the stability of MRN subunits, leading to the increased kinase activity of ATM in response to DNA damage (Figure 3.20).

Figure 3.20: A novel possible role of FOXM1 in DNA damage response pathway (red). FOXM1 regulates NBS1 expression and ATM kinase activity in response to DNA damage.

As NBS1 has been implicated in HR repair via the formation of MRN complex (Yuan and Chen, 2010), I then asked whether FOXM1 and NBS1 function together in DSB repair via HR using the HeLa cells containing a DR-GFP reporter construct. As expected, the depletion of either FOXM1 or NBS1 leads to a significant reduction in

HR. By contrast, overexpression of FOXM1 or NBS1 enhanced the HR repair. These data confirmed that FOXM1 and NBS1 play an important role in HR repair. To explore the functional link between FOXM1 and NBS1 in HR pathway, HeLa cells carrying DR-GFP were co-transfected with FOXM1 plasmid and NBS1 siRNA. The results revealed that overexpression of FOXM1 together with depletion of NBS1 lead to a strong decrease in HR repair ability of the HeLa cells (Figure 3.12). These data

144 imply that FOXM1 is a crucial regulator of the DNA repair gene, NBS1, and both proteins are required for proper DNA DSB repair responses. In agreement, the ectopic expression of NBS1 was able to rescue the DNA repair defects and epirubicin-induced senescence phenotypes in Foxm1-/- MEFs cells. Taken together, these findings clearly confirm that the role of FOXM1 in HR-mediated DSB repair is linked to its ability to control NBS1 expression. Consistent with my result, NBS1 has been shown previously to play a role in persistent DNA damage-induced senescence-associated inflammatory cytokine secretion (Rodier et al., 2009a).

I next tested the potential of targeting NBS1 to improve the effectiveness of epirubicin treatment in both sensitive and resistant MCF-7 breast cancer cells by using SRB assays. The result showed that downregulation of NBS1 rendered MCF-7

EpiR cells more sensitive to epirubicin, suggesting that NBS1 is a novel modulator of epirubicin sensitivity (Figure 3.14). The enhanced epirubicin sensitivity of the NBS1 depleted MCF-7 EpiR cells were further confirmed by the increased expression of cleaved PARP protein (Figure 3.15). As FOXM1 is involved in DNA damage-induced senescence after long-term treatments with low concentrations of epirubicin and also controls NBS1 expression, I then examined the effects on NBS1 depletion on the long-term proliferation. Like FOXM1, NBS1-knockdown inhibited long-term cell proliferation and triggered premature senescence in MCF-7 and MCF-7 EpiR cells in response to low doses of epirubicin treatment (Figure 3.16A and 16C, respectively).

Collectively, these findings strongly suggest that NBS1 protects MCF-7 and MCF-7

EpiR cells from entering senescence/ cell death and it could be used as a potential target to improve the effectiveness of epirubicin treatments in the resistant cancer cells. In agreement, expression of an NBS1 mutant, which interrupts the MRN

145 function for DNA repair, has been shown to be associated with an increased sensitivity to cisplatin in head and neck cancer (Tran et al., 2004, Araki et al., 2010).

The physiological relevance of the regulation of NBS1 by FOXM1 is further underscored by the significant correlation between nuclear FOXM1 and total NBS1 expression in breast cancer patient samples (Figure 3.18). Furthermore, both

FOXM1 and NBS1 expression is significantly correlated to poor prognosis in breast cancer, supporting a physiological role of FOXM1 and NBS1 in genotoxic drug resistance.

In summary, this study has identified NBS1 as an important novel target of FOXM1 involved in HR-mediated DSB repair, DNA damage-induced senescence and epirubicin resistance. These findings also show the potential therapeutic benefit of

FOXM1-NBS1 axis as a reliable prognostic marker for monitoring treatment efficiency and a promising target for therapeutic intervention to overcome epirubicin resistance in breast cancer.

146

CHAPTER 4

THE ROLE OF FOXM1 IN MITOTIC CONTROL

AND PACLITAXEL RESISTANCE

147

4.1 Introduction

Paclitaxel represents one of the most widely used mitotic poison agents for treating advanced breast cancer. At the molecular level, paclitaxel bind to the β-tubulin subunit of microtubules, promoting the polymerisation of tubulin resulting in the formation of microtubule asters in interphase cells, instead of mitotic spindles, during mitosis (Bedard et al., 2010). This process blocks cell cycle progression through the

M-phase by interrupting normal microtubule dynamics and inducing cancer cell death

(Jiang et al., 2011). Apoptosis has generally been thought to be the most predominant mechanism of cell death in response to taxanes chemotherapy at the high drug doses (Chang et al., 1999, Roninson et al., 2001). However, in this chapter, I hypothesised that other modes of cell death, including treatment-induced senescence and mitotic catastrophe may also contribute significantly to the overall therapeutic response.

Mitotic catastrophe is characterised by the occurrence of aberrant mitosis, or mis- segregation of the chromosomes, followed by cell division. Nuclear envelopes form around individual chromosomes or groups of chromosomes forming large cells with multiple micronuclei, which are morphologically distinguishable from apoptotic cells

(Morse et al., 2005).

Despite clinical effectiveness of paclitaxel, toxic side effects and the development of drug resistance represent major obstacles to improve outcome for breast cancer patients. Therefore, understanding the molecular mechanisms of drug resistance and developing agents to reverse the resistance are pivotal for improving the prognosis of cancer patients.

148

Studies on breast cancer have revealed that overexpression of FOXM1 also confers resistance to paclitaxel, whereas depletion of FOXM1 can increase sensitivity of breast cancer cell to this drug (Carr et al., 2010). However, the exact mechanisms by which FOXM1 modulates paclitaxel resistance remain elusive. The possible processes could be related to the specific relationship between FOXM1 and co- factors that are involved in spindle assembly and mitotic control. The mitotic kinesin

KIF20A has been reported to be a potential downstream FOXM1 target, which are involved in mitosis and chromosome segregation (Wonsey and Follettie, 2005).

Consistently, our microarray analysis result in the laboratory also revealed that

FOXM1 might regulate the expression of KIF20A in response to paclitaxel treatment in MCF-7 cells. KIF20A, also known as RAB6KIFL/MKLP2, is a member of the kinesin family of microtubule motor proteins which plays many important roles in cellular functions including mitotic spindle assembly, cytokinesis, cell migration as well as localisation of organelles and vesicles (Taniuchi, Nakagawa et al. 2005, Hill et al., 2000, Gruneberg et al., 2004, Neef et al., 2006). KIF20A expression appears to be tissue specific. It is widely expressed in human fetal tissues. It is also expressed in the adult during haematopoiesis and in various proliferating tissues, including thymus, bone marrow and testis (Lai et al., 2000). By contrast, there is no expression of KIF20A in adult quiescent human liver cells, suggesting that KIF20A contributes to both normal and pathological proliferation as well as cancer progressiveness in human cells (Gasnereau et al., 2012). Interestingly, accumulating evidence has shown that KIF20A is overexpressed in pancreatic ductal adenocarcinoma (PDAC) cells and its downregulation inhibits cell growth, migration and invasion (Taniuchi, Nakagawa et al. 2005, Yan et al., 2012, Stangel et al., 2015,

Exertier et al., 2013). Recently, a KIF20A-derived peptide vaccine KIF20A-66 has

149 been developed and used as a promising and effective immunotherapy to treat advanced pancreatic cancer (Asahara et al., 2013, Suzuki et al., 2014). Moreover, a growing number of studies have consistently revealed that KIF20A is also abundantly overexpressed in many other types of cancers, such as gastric cancer, small lung cancer, melanoma and bladder cancer (Yan et al., 2012, Yamashita et al.,

2012, Ho et al., 2012). However, little is known about the transcriptional regulation of

KIF20A and it remains to be elucidated whether KIF20A plays a role in carcinogenesis or anti-cancer drug resistance in breast cancer cells.

In this chapter, I hypothesise that FOXM1 might regulate the mitotic regulatory protein KIF20A and this ultimately determines the paclitaxel-sensitivity or resistance in breast cancer cells. Therefore, the aim of this chapter is to study the functional relationship between FOXM1 and KIF20A in mediating paclitaxel sensitivity and resistance in breast cancer.

150

4.2 Results

4.2.1 Deletion of FOXM1 reduces long-term clonogenic survival and induces cellular senescence in response to paclitaxel treatment

Studies in our laboratory and others have shown that FOXM1 overexpression is sufficient to render breast cancer cells resistant to a wide range of anticancer chemotherapy agents, including epirubicin, lapatinib, gefitinib, imatinib and cisplatin

(Myatt et al., 2014, Lam et al., 2013). In this chapter, I investigated the possibility that upregulation of FOXM1 in cancer cells could confer resistance to paclitaxel. To assess the potential requirement of FOXM1 in response to paclitaxel treatment, I initially treated early passage WT and Foxm1-/- MEFs with a wide range of concentrations of paclitaxel (0, 20, 40, 60, 80 and 100 nM). Long-term cellular viability was evaluated by clonogenic assay. The results revealed that Foxm1-/- MEFs displayed much higher sensitivity to paclitaxel compared to the WT MEFs (Figure

4.1A). In order to determine whether this long-term loss of proliferative capacity was due to senescence, the WT and Foxm1-/- MEFs were subjected to SA-βgal staining to identify senescent cells. Consistent with the results obtained in the clonogenic assays, FOXM1 deletion in MEFs significantly induced cellular senescence in response to paclitaxel treatment, as evidenced by the increase in SA-βgal activity and flat cell morphology (Figure 4.1B).

151

Figure 4.1: Deletion of FOXM1 reduces long-term clonogenic survival and induces cellular senescence in response to paclitaxel treatment in MEFs. (A) Clonogenic assay was performed to determine the colony formation efficiency of Foxm1−/− and WT MEFs. 2,000 cells were seeded in six well plates, treated with 0, 20, 40, 60, 80 and 100 nM of paclitaxel and grown for 15 days. The cells were then stained with crystal violet. The bar graph represents average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01). (B) SA-βgal staining of Foxm1−/− and WT MEFs treated with 0, 20, 40, 60, 80 and 100 nM of paclitaxel. 5 days after treatment, cells were stained for SA-β-galactosidase activities.

152

4.2.2 Knockdown of FOXM1 suppresses cell proliferation in both paclitaxel sensitive and resistance MCF-7 cells

To confirm the role of FOXM1 in breast cancer progression and paclitaxel resistance,

MCF-7 and the paclitaxel resistant MCF-7 TaxR breast cancer cells were transfected with control siRNA or siRNA targeting FOXM1. The effect of FOXM1 depletion on colony formation was next evaluated by clonogenic assay. The results showed that

FOXM1 knockdown was able to sensitise MCF-7 cells to paclitaxel at the very low concentrations of paclitaxel and there was no difference in cell viability between the

MCF-7 cells treated with 3, 5 or 10 nM paclitaxel (Figure 4.2A). Interestingly, FOXM1 depletion alone significantly decreased the colony forming ability of MCF-7 TaxR cells which did not depend on the dosage of paclitaxel used suggesting that FOXM1 can protect cells from paclitaxel-induced cell death (Figure 4.2B).

153

Figure 4.2: FOXM1 knockdown suppresses the clonogenic ability of MCF-7 and MCF-7 TaxR cells. A) MCF-7 and (B) MCF-7 TaxR were transfected with NS siRNA or FOXM1 siRNA. 24 h after transfection, 2,000 cells were seeded in six well plates, treated with 0, 1, 3, 5, 7.5 and 10 nM of paclitaxel, grown for 15 days and then stained with crystal violet. The results represent average of two independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (* P ≤ 0.05, ** P ≤ 0.01 and n.s. for non-significant).

154

4.2.3 Elevated expression of FOXM1 confers paclitaxel resistance.

In order to test if overexpression of FOXM1 is directly responsible for conferring paclitaxel resistance, I performed transient transfection with either pcDNA3 empty vector or pcDNA3-FOXM1 plasmids in MCF-7 cells, and evaluated cell viability using the sulforhodamine-B (SRB) assay. The result showed that the overexpression of

FOXM1 was sufficient to increase paclitaxel resistance to the parental MCF-7 cells

(Figure 4.3). These data suggest that overexpression of FOXM1 reduces sensitivity to paclitaxel by protecting MCF-7 cells from undergoing proliferative arrest.

Figure 4.3: Overexpression of FOXM1 induces resistance to paclitaxel in MCF-7 cells. MCF-7 cells were transfected with control empty vector or pcDNA3-FOXM1 plasmid. Cells were then treated with increasing concentrations of paclitaxel (1 – 50 nM) for 48 h and their cell viability measured by sulforhodamine B (SRB) assay. Representative data from three independent experiments are shown. Statistical analyses were performed using Student’s t-tests (**P≤0.05).

155

4.2.4 KIF20A and FOXM1 are upregulated in paclitaxel resistant MCF-7 (MCF-7

TaxR) cells

The mitotic kinesin KIF20A has been reported to be a potential downstream FOXM1 target, which are involved in mitosis and chromosome segregation (Wonsey and

Follettie, 2005). To elucidate the possible functional role of FOXM1 in paclitaxel resistance and the mechanism of action involved, I determined the protein expression levels of FOXM1 and its putative target KIF20A upon the paclitaxel treatment in both sensitive MCF-7 cells and paclitaxel resistant MCF-7 TaxR cells.

The western blot analysis revealed that FOXM1 expression greatly decreased in the sensitive MCF-7 cells in response to the moderate dose of paclitaxel (10 nM), while the expression of FOXM1 in MCF-7 TaxR cells were maintained at high levels upon the treatment. Interestingly, the protein expression levels of KIF20A displayed a similar kinetic pattern observed for FOXM1 upon the paclitaxel treatment in both cell lines, indicating that FOXM1 possibly mediates paclitaxel sensitivity through regulation of KIF20A (Figure 4.4A, Figure S6). Interestingly, there was no change in the expression of PARP in MCF-7 cells, suggesting that other alternative forms of programmed cell death, such as senescence or mitotic catastrophe, can be triggered in response to 10 nM paclitaxel treatment. Consistently, RT-qPCR analysis showed that FOXM1 and KIF20A mRNA levels were significantly upregulated in MCF-7 TaxR cells compared with the sensitive MCF-7 cells which exhibited decreased KIF20A transcriptional levels by almost 50% after 24 h paclitaxel treatment (Figure 4.4B).

However, the reduction in mRNA levels of FOXM1 and KIF20A were observed at

48h paclitaxel treatment of MCF-7 TaxR cells. It is likely to be due to the MCF-7 TaxR cells becoming over-confluent as these cells proliferate rapidly and are not inhibited by paclitaxel causing the cells to undergo either cell-cycle arrest or cell death. Taken

156 together, these results suggest that both KIF20A and FOXM1 proteins might be mediators of paclitaxel resistance in MCF-7 breast cancer cells.

Figure 4.4: FOXM1 and KIF20A expression levels were strongly upregulated in paclitaxel- resistant MCF-7 cells. (A) Protein expression levels of FOXM1 , KIF20A, cyclin B1 and PARP in MCF-7 and MCF-7 TaxR cell lines following paclitaxel treatment at different time points were determined by western blot analysis (left panel). RT-qPCR analysis determining the relative mRNA expression levels of FOXM1 and KIF20A in MCF-7 and MCF-7 TaxR cell lines following paclitaxel treatment at different time points. The results represent average of three independent experiments ± SD. Statistical significance was determined by Student’s t-test, **P ≤ 0.01.

157

4.2.5 Silencing of FOXM1 downregulates KIF20A expression

According to the previous western blot analysis (Figure 4.4), FOXM1 and KIF20A expression are increased in MCF-7 TaxR cells compared to sensitive MCF-7 cells. To test if KIF20A is a downstream target of FOXM1, the expression of KIF20A was examined by RT-qPCR and western blot analysis following silencing FOXM1 using siRNA in MCF-7 and MCF-7 TaxR cells. The results showed that depletion of FOXM1 significantly reduced the expression of KIF20A at both mRNA and protein levels

(Figure 4.5A, Figure 4.5B and Figure S7), indicating that FOXM1 regulates expression of KIF20A. To confirm this further, the protein expression levels of

FOXM1 and KIF20A were investigated by western blot analysis in WT and Foxm1-/-

MEFs. I observed that in the Foxm1-/- MEFs there was a decreased expression of

KIF20A (Figure 4.5C). Conversely, ectopic overexpression of FOXM1 in MCF-7 breast cancer cells induced the expression of KIF20A (Figure 4.5D). In agreement, the KIF20A mRNA and protein were also downregulated after FOXM1 depletion in

MDA-MB-231 cells (Figure 4.6).

158

Figure 4.5: Silencing of FOXM1 downregulates KIF20A expression in MCF-7 and MCF-7 TaxR cells. MCF-7 and MCF-7 TaxR cells depleted in FOXM1 expression revealed significant decreases in both mRNA (A) and protein (B) levels of KIF20A expression. Bars represent mean± S.D. Statistical significance was determined by Student’s t-test, (**P≤0.01, significant). (C) Foxm1−/− MEFs showed decreased protein expression of KIF20A. (D) Overexpression of FOXM1 caused an induction KIF20A expression levels in MCF-7 cells.

159

Figure 4.6: Depletion of FOXM1 causes a decrease in the expression of KIF20A in MDA-MB-231 cells. MDA-MB-231 cells were transfected either with FOXM1 siRNA or non-silencing controls (NS). (A) FOXM1 and KIF20A mRNA levels were determined by RT-qPCR. Bars represent mean± S.D. Statistical significance was determined by Student’s t-test, (**P≤0.01, significant). (B) The expression levels of FOXM1, KIF20A and β-tubulin were analysed by western blot analysis.

160

4.2.6 Overexpression of FOXM1 enhances the transcriptional activity of KIF20A promoter

It has been reported that FOXM1 binds to promoter regions of its targets with a preference for tandem repeats of a core consensus 'A(T/C)AAA(T/C)AA' recognition sequence (Myatt and Lam, 2007, Littler et al., 2010). In order to determine whether

KIF20A is a direct downstream target of FOXM1, I initially performed luciferase assay in MCF-7 cells by co-transfection the KIF20A promoter luciferase reporter construct, previously generated in the laboratory by my college Dr Fung Zhao, with the FOXM1B-pcDNA3 plasmid. This KIF20A promoter construct contained a 1.1 kbp of region (−1150/−61) upstream of the most 5’-transcription start site. Five possible

FOXM1 consensus binding sites were identified in this region and one of them is located at -91, as described in Figure S9. However, the luciferase reporter assay result revealed that FOXM1 was unable to transactivate and induce KIF20A transcription activity via this 5’-UTR region. Thus, I next analysed the consensus

FOXM1 binding sequence in other region of KIF20A gene using the MCF-7 ChIP-

Seq data (hg19: GSM1010769) from the Encyclopedia of DNA Elements (ENCODE) project (Landt et al., 2012) and identified strong FOXM1 occupancy at a region (-

21/+144) mapped downstream of the most 5’-transcription start site but upstream of the 2nd transcription start site (Figure 4.7A). From sequence analysis, a putative forkhead responsive element (FHRE) was identified within (+80bp regions) (Figure

4.7A). To test whether this FOXM1 binding site in the KIF20A sequence (+80bp) was functional, I generated three new reporter plasmids: pGL3-KIF20A-WT, pGL3-

KIF20A-MUT1 (single site mutation at +80) and pGL3-KIF20A-MUT2 (double site mutation at -91 and +80 to test the specificity of FOXM1 binding site at +80) (Figure

4.7A). To construct these three plasmids, the KIF20A-WT, KIF20A-MUT1, and

161

KIF20A-MUT2 fragments (Figure S10) were initially designed and synthesised by

GeneArt (Life Technologies, Darmstadt, Germany). The synthesised DNA fragments were then amplified and cleaved with the restriction enzymes XhoI and Bglll and cloned into the XhoI/Bglll sites of the pGL3-basic vector (Promega, Madison, WI).

DNA extracted from all the positive clones were re-digested with XhoI and Bglll enzymes and performed gel electrophoresis to confirm the gene inserts. As seen in the Figure 4.7B, the obtained bands correlated with the size of insert fragments (384 bp) and pGL3-basic empty vector (4818 bp). These results indicate that the fragments were successfully cloned into the pGL3-basic luciferase reporter vector.

I next assayed the ability of FOXM1 to induce KIF20A expression using a transient luciferase reporter assays in MCF-7 cells. As shown in the Figure 4.7C, the KIF20A-

WT promoter activity displayed a significant induction in response to ectopic FOXM1 overexpression, whereas this significant induction was not observed in cells transfected with either of the mutant constructs of KIF20A promoter (MUT1 and

MUT2). Interestingly, there was no further significant reduction in KIF20A promoter activity observed in the cells transfected with double mutation at both -91 and 80+bp

(MUT2) construct, when compared to the cells transfected with single mutation at

+80 bp (MUT1) construct. This result suggests that FOXM1 is able to activate

KIF20A transcription activity specifically through the consensus FOXM1 binding site located at +80 bp.

4.2.7 FOXM1 directly binds to KIF20A loci in MCF-7 cells.

To confirm further that FOXM1 directly binds to KIF20A promoter region at +80 bp, I performed ChIP analysis in MCF-7 cells using specific primers (+8/+131) to amplify the putative FOXM1 binding site (Figure 4.8A). The chromatin fragments from empty vector control and FOXM1-overexpressing MCF-7 cells were immunoprecipitated

162 with either FOXM1 or IgG negative control antibodies. The binding of FOXM1 at

KIF20A sequence was quantified by RT-qPCR analysis. As shown in Figure 4.8B

(top-left panel), the ChIP analysis showed that overexpression of FOXM1 enhances the enrichment of the binding of FOXM1 to the KIF20A promoter region. In contrast, the inhibition of FOXM1 expression and transcription activity by treating with 10 µM thiostrepton, as previously described (Kwok et al., 2008), significantly decreases the

FOXM1 occupancy (Figure 4.8B, bottom-left panel), suggesting that FOXM1 is able to bind and transactivate the KIF20A gene through the FHRE located at position

(+80 bp). Taken together, these results strongly support the notion that FOXM1 is a crucial transcriptional regulator of the KIF20A.

163

Figure 4.7: Overexpression of FOXM1 enhances the transcriptional activity of KIF20A promoter in MCF-7 cells. (A) Schematic representation of the human KIF20A promoter region depicting the FOXM1 binding site (wild-type (+80)) and the FOXM1 binding site-mutated MUT1 and MUT2. (B) Cloning of 384 bp KIF20A promoter fragment into pGL3basic. pGL3 basic harbouring either WT, MUT1 or MUT2 promoter was re-digested with XhoI and Bglll and ran on a 1% agarose gel to confirm the constructs. (C) MCF-7 cells were transiently co-transfected with wild-type or FOXM1 binding site- mutated (mut) KIF20A reporter plasmids and Renilla luciferase plasmid along with or without FOXM1 expression plasmids. Twenty-four hours after transfection, cells were lysed, and the luciferase activity was examined. Firefly luminescence signal was normalised based on the Renilla luminescence signal. The results represent average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (** P ≤ 0.01 and n.s. for non-significant).

164

Figure 4.8: FOXM1 directly binds to the KIF20A promoter region in MCF-7 cells. (A) Schematic representation of the human KIF20A promoter region depicting the FOXM1 binding site (FHRE) and the ChIP-qPCR specific primer regions: to amplify the putative FOXM1 binding site (+8/+131) and negative control site located at the 5’ upstream non-coding region (-813/-938 bp) of KIF20A gene(B) The binding of FOXM1 to the KIF20A promoter was determined by ChIP analysis in FOXM1- overexpressing (B, upper panel) or FOXM1-depleted (B, lower panel) MCF-7 cells. The chromatin fragments obtained from transfected cells were immunoprecipitated with either FOXM1 or IgG negative control antibodies. The binding of FOXM1 at KIF20A sequence was then quantified by RT- qPCR analysis. The results represent average of two independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (** P ≤ 0.01).

165

4.2.8 Low concentrations of paclitaxel cause aberrant mitosis in MCF-7 cells

Apoptosis has been considered as the main mechanism of cell death in response to chemotherapy (Ricci and Zong, 2006). Currently, a more heterogeneous model of therapeutic response is acknowledged wherein multiple forms of death combine to generate the overall tumour response. The resulting mechanisms of cell death are determined by several factors, including the mechanism of action of the therapy, the dose regimen used, and the genetic background of the cells within the tumour

(Morse et al., 2005). Paclitaxel-mediated cell death is usually studied at high concentrations of drug (10 to 200 nM) that cause mitotic arrest and rapid cell death

(Blagosklonny and Fojo, 1999, Zasadil et al., 2014). In this study, I tested whether low concentrations of paclitaxel (≤5 nM) are sufficient to render MCF-7 cells to undergo cell death. To determine the cellular consequences of treatments with low concentrations of paclitaxel, MCF-7 cells were incubated with the indicated doses of paclitaxel and the mitotic spindle formation was then examined using α-tubulin antibodies to stain microtubules, γ-tubulin to stain centrosome and DAPI to stain

DNA. Figure 4.9A (top panel) indicates a control (untreated) cell in mitosis, with a typical bipolar spindle. Figure 4.9A (lower panels) represent various mitotic abnormalities, including bipolar spindles with an irregular chromosome alignment, monopolar and multipolar spindles, observed in paclitaxel-treated MCF-7 cells.

Quantitative analysis of mitotic cells stained with α-tubulin and γ-tubulin antibodies

(Figure 4.9B) indicates that ~80% of 5 nM paclitaxel treated MCF-7 cells exhibited mitotic spindle defects.

166

167

Figure 4.9: Low concentrations of paclitaxel cause mitotic catastrophe. MCF-7 cells were treated with or without 5 nM paclitaxel for 24 h. Cells were then immunostained with α-tubulin antibody (Green) and γ-tubulin antibody (Red). Nuclei were stained with DAPI (Blue). Metaphase cells were captured using Leica TCS SP5 (63X magnification). For each condition, images of at least 50 mitotic cells were analysed. (A) Representative confocal images are shown. The metaphase cells classified into four distinct spindle formation types, including normal bipolar, bipolar with unaligned- chromosome, monopolar or multipolar spindles, were quantified. (B) Bars represent average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (* P ≤ 0.05, **P≤0.01, significant; n.s. for non-significant).

168

4.2.9 Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation and chromosome alignment.

To address how the lack of FOXM1 affects mitotic progression in MCF-7 cells, the subcellular distribution of α-tubulin and chromosomes was examined in metaphase of MCF-7 cells after FOXM1 and KIF20A knockdown (Figure 4.10). In the control siRNA-transfected cells, condensed chromosomes were aligned properly at the metaphase plate with bipolar spindles. By contrast, the FOXM1 and KIF20A- depleted cells displayed abnormal mitotic spindles, including monopolar and multipolar mitotic spindles as well as bipolar spindles with unaligned chromosomes.

The frequency of abnormal mitotic spindles in MCF-7 cells was increased approximately 3-fold compared to the control by FOXM1 knockdown and about 4- fold by KIF20A knockdown (Figure 4.10A-C). The failure to establish normal mitotic spindles in metaphase induced by KIF20A and FOXM1 depletion also caused a significant increase in lagging chromosomes in anaphase (Figure 4.11) and ultimately, the accumulation of large multinucleated cells, indicative of mitotic catastrophe (Figure 4.12). These results indicated that FOXM1 and KIF20A are essential for the formation of normal mitotic spindles, and defects of which lead to abnormal chromosome segregation and mitotic progression. Interestingly, in the paclitaxel-treated MCF-7 cells the increase in abnormal spindle formation after

FOXM1 or KIF20A silencing was no longer apparent. Together with my previous results, these findings suggest that FOXM1 and KIF20A modulate the cytostatic and cytotoxic function of paclitaxel through regulation of mitotic spindle formation.

169

170

Figure 4.10: Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation and chromosome alignment. (A and B) MCF-7 cells were transfected with NS, FOXM1 or KIF20A siRNA. 24 h after transfection, cells were treated with 5 nM paclitaxel for 24 h. Cells were then immunostained with α-tubulin antibody (Green). Nuclei were stained with DAPI (Blue). Metaphase cells were visualised with Leica TCS SP5 (63X magnification). For each condition, images of at least 50 mitotic cells were captured. Representative confocal images are shown. (C) The number of metaphase cells classified into either normal bipolar, abnormal bipolar, monopolar or multipolar spindles were quantified. Results represent average of three independent experiments ± S.D.

171

Figure 4.11: Depletion of FOXM1 or KIF20A causes anaphase bridges, lagging chromosomes, and chromosome instability in MCF-7. MCF-7 cells were treated with or without 5 nM paclitaxel for 24 h. immunostained with α-tubulin antibody (Green) and γ-tubulin antibody (Red). Nuclei were stained with DAPI (Blue). Metaphase cells were captured using Leica TCS SP5 (63X magnification). Mitotic cells were visualised with Leica TCS SP5 (63X magnification).For each condition, images of at least 50 anaphase cells were captured to analyse for the frequency of anaphase bridges and/or lagging chromosome. Representative confocal images are shown. Results represent average of three independent experiments ± S.D. Yellow arrows indicate chromosomal abnormalities, such as lagging chromosomes and chromosome bridges.

172

35 µm

Figure 4.12: Characterisation of mitotic catastrophe in MCF-7 cells following FOXM1 or KIF20A knockdown or paclitaxel treatment. MCF-7 cells were transfected with NS, FOXM1 or KIF20A siRNA. 24 h after transfection, cells either left untreated or treated with 5 nM paclitaxel for 24 h. Cells were then immunostained with antibody against α-tubulin (Green). Nuclei were stained with DAPI (Blue). Cells were visualised with Leica TCS SP5 (63X magnification).

173

4.2.10 Intracellular localisation of FOXM1 and KIF20A during mitosis

To study the functional significance of FOXM1 and KIF20A in mitosis, the subcellular localisation of FOXM1 and KIF20A were examined at different stages of mitosis in

MCF-7 cells by immunofluorescent staining with FOXM1 and KIF20A antibodies

(Figure 4.13). During interphase, the staining pattern of FOXM1 in the nucleus overlapped with KIF20A. However, KIF20A was also detected in the cytoplasm, indicating that in the interphase cells, KIF20A participates in retrograde vesicular traffic between the Golgi apparatus and the endoplasmic reticulum (Gasnereau et al., 2012). In metaphase cells possessing a single spindle, the two proteins co- localised in cytoplasm. At anaphase, FOXM1 and KIF20A staining accumulated in the midzone of the spindle. When cells were at telophase, both FOXM1 and KIF20A were present sharply concentrated in the midbody, further suggesting that FOXM1 and KIF20A may function together and are essential for cell cycle regulation during successful cytokinesis.

174

Figure 4.13: Dynamic localisation of FOXM1 and KIF20A during mitosis of MCF-7 cells. Immunofluorescence analysis of FOXM1 and KIF20A localisation during mitosis of MCF-7 cells by labelling with antibodies against FOXM1 (Green) and KIF20A (Red). Nuclei were counterstained with DAPI (Blue). Stained mitotic cells were visualised with Leica TCS SP5 (63X magnification).

175

4.2.11 Depletion of KIF20A or FOXM1 reduces long-term clonogenic survival and induces cellular senescence in MCF-7 and MCF-7 TaxR cell lines.

I next tested the efficiency of targeting FOXM1 and KIF20A in MCF-7 and the paclitaxel-resistant MCF7-TaxR breast cancer cell lines. According to previous long term clonogenic assay results in Figure 4.2, it showed that there was no difference in cell viability between the MCF-7 cells treated with 3, 5 or 10nM paclitaxel. Therefore, in this experiment, cells were transfected with siRNA targeting FOXM1 and KIF20A and long-term cell viability rates were evaluated by clonogenic assay at very low doses of paclitaxel (1 and 3 nM). The results showed that knockdown of FOXM1 sensitised MCF-7 cells to long-term proliferative arrest at relatively low paclitaxel dose. Similar to FOXM1 depletion, knockdown of KIF20A also increased the sensitivity of MCF-7 cells to paclitaxel at the same concentration of the drug (Figure

4.15A). Interestingly, depletion of FOXM1 or KIF20A alone almost completely inhibited the colony-forming capacity of MCF-7 TaxR cells irrespective of the dosage of paclitaxel used, suggesting that MCF-7 TaxR cells are dependent on high expression levels of FOXM1 and KIF20A for long-term survival (Figure 4.17A).

Consistent with the clonogenic assay results, FOXM1 and KIF20A knockdown in

MCF-7 TaxR cells significantly induced senescence-associated SA-βgal activity and morphology independent of the paclitaxel concentration, suggesting that MCF-7

TaxR cells have become dependent on FOXM1 and KIF20A expression to override the non-apoptotic pathway including cellular senescence (Figure 4.17B).

Transfection efficiency of FOXM1 siRNA and KIF20A siRNA knockdown in MCF-7

(Figure 4.14) and MCF-7 TaxR (Figure 4.16) was confirmed by monitoring the uptake of the BLOCK-iT fluorescent oligo (green fluorescence) at 48 h following transfection using Oligofectamine. The results showed that more than 80% of cells were

176 successfully transfected. The silencing effect was also tested by RT-qPCR and western blot analysis.

Figure 4.14: The transfection efficiency of FOXM1-siRNA and KIF20A-siRNA in MCF-7 cells. (A)Transfection efficiency was monitored by the uptake of the BLOCK-iT fluorescent oligo (green fluorescence) at 48 h following transfection using Oligofectamine. More than 70% of cells were successfully transfected. (B) FOXM1 and KIF20A mRNA expression levels were detected by real-time qPCR at 48 h post-transfection. Bars represent average±S.D. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P≤0.01). (C) FOXM1 and KIF20A protein expression levels were detected by Western blot analysis at 48 h post-transfection.

177

Figure 4.15: Targeting KIF20A or FOXM1 using siRNA inhibits cell growth and induces senescence in MCF-7 cells. (A) MCF-7 cells were transfected with either NS (non-targeting control) siRNA, siRNA targeting FOXM1 or KIF20A. 24 h after transfection, 2,000 cells were seeded in six well plates, treated with 0, 1, or 3 nM of paclitaxel, grown for 15 days and then stained with crystal violet (left panel). The result (right panel) represents average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P 0.05, **P 0.01, ***P 0.005; n.s., non-significant). In parallel, (B) MCF-7 transfected with NS, FOXM1 or KIF20A siRNA were seeded in six-well plates, treated with 0, 1, or 3 nM of paclitaxel. Five days after treatment, cells were stained for SA-βgal activity. The graph shows the percentage of SAβ-gal-positive cells as measured from five different fields from two independent experiments. Bars represent average±S.D. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P≤0.05, **P≤0.01, ***P≤0.005, significant; n.s., non-significant).

178

Figure 4.16: The transfection efficiency of FOXM1-siRNA and KIF20A-siRNA in MCF-7 TaxR cells. (A)Transfection efficiency was monitored by the uptake of the BLOCK-iT fluorescent oligo (green fluorescence) at 48 h following transfection using Oligofectamine. More than 70% of cells were successfully transfected. (B) FOXM1 and KIF20A mRNA expression levels were detected by real-time qPCR at 48 h post-transfection. Bars represent average± S.D. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P≤0.01). (C) FOXM1 and KIF20A protein expression levels were detected by Western blot analysis at 48 h post-transfection.

179

Figure 4.17: Targeting KIF20A or FOXM1 using siRNA significantly impairs colony formation and induces senescence in MCF-7 TaxR cells. (A) MCF-7 TaxR cells were transfected with either NS, FOXM1 siRNA or KIF20A siRNA. 24 h after transfection, 2,000 cells were seeded in six well plates, treated with 0, 1, or 3 nM of paclitaxel, grown for 15 days and then stained with crystal violet (left panel). The result (right panel) represents average of three independent experiments ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P 0.01, significant; n.s., non-significant). In parallel, (B) MCF-7 TaxR transfected with NS, FOXM1 or KIF20A siRNA were seeded in six-well plates, treated with 0, 1, 3 nM of paclitaxel. Five days after treatment, cells were stained for SA-βgal activity. The graph shows the percentage of SAβ-gal-positive cells as measured from five different fields from two independent experiments. Bars represent average±S.D. Statistical significance was determined by two-tailed unpaired Student’s t-test (*P≤0.05, **P≤0.01, significant; n.s., non-significant).

180

4.2.12 Correlation between FOXM1 and KIF20A expression in breast cancer samples

After establishing that KIF20A is a direct transcriptional target of FOXM1 in breast cancer, in collaboration with Professor Khoo Ui Soon (Department of Pathology, the

University of Hong Kong), we established further the physiological significance of the regulation of KIF20A by FOXM1 in breast cancer using the same set of TMA described in chapter 3 (n=116), but only 100 cases had assessable KIF20A scores.

Tissues of the remaining 16 cases were missing. For the immunohistochemical analysis, KIF20A score was dichotomized using the median of all the KIF20A scores.

So half of the cases were assigned as KIF20A high and half were assigned as

KIF20A low. The results revealed that FOXM1 and KIF20A proteins were detected in both nuclear and cytoplasmic compartments. Statistical analysis of the expression patterns of these proteins showed that there was a strongly significant correlation between FOXM1 and KIF20A expression (Pearson coefficient r=0.292, P=0.006 for total KIF20A; r=0.250, P=0.019 for cytoplasmic KIF20A; r=0.228, P=0.034 for nuclear KIF20A) (Figure 4.18A and 4.18B). This further supported my finding in the cell lines that FOXM1 directly regulates KIF20A transcription. In addition, survival analysis by Kaplan-Meier estimate with log-rank test revealed that overexpression of nuclear KIF20A was significantly associated with poor prognosis (P = 0.045 for overall survival and P=0.016 for disease-specific survival, respectively) (Figure

4.19A). In multivariate Cox regression analysis, KIF20A nuclear staining remained associated with poor survival after correcting for tumour stage and lymph-node involvement (P = 0.047, RR=2.47 for overall survival and P=0.037, RR=2.767 for disease-specific survival, respectively) (Figure 4.20A), suggesting that KIF20A nuclear score is a prognostic marker independent of the clinicopathological

181 parameters examined. In this cohort, 60% of patients received chemotherapy. For these patients, high nuclear KIF20A expression was significantly associated with poor survival (log-rank test, P=0.008 for overall survival and P=0.004 for disease- specific survival, respectively) (Figure 4.19B) and presented a stronger risk marker

(P=0.013, RR=4.008 for overall survival and P=0.01, RR=5.089 for disease-specific survival, respectively) (Figure 4.20B), suggesting that similar to FOXM1, KIF20A expression is associated with chemotherapeutic drug resistance. In agreement, further Kaplan-Meier analysis of KIF20A and FOXM1 mRNA transcript expression in a previously published cohort (1,809 breast cancer patients) (Györffy et al., 2010) confirmed that both high FOXM1 and KIF20A mRNA expression levels are very significantly related to poor survival (p<0.00001 and p<0.00001, respectively for overall survival) (Figure 4.19C). The significance of both FOXM1 and KIF20A presented in different survival analyses further strengthened the crucial involvement of both genes in breast cancer progression and chemotherapeutic drug response.

182

Figure 4.18: Correlation between FOXM1 and KIF20A expression in breast cancer samples. (A) FOXM1 and KIF20A expression was assessed by immunohistochemistry using tissue-microarray (TMA) slides containing 100 breast cancer patient samples. KIF20A score is dichotomized using the median of all the KIF20A scores. So half of the cases are assigned as KIF20A high and half are assigned as KIF20A low. Representative staining images of one patient with high FOXM1 and KIF20A expression and one with low expression were shown. Positive correlation between FOXM1 and KIF20A was observed. (B) KIF20A staining was detected in both nuclear and cytoplasmic compartments and was correlated with FOXM1 staining. Statistical analysis revealed that KIF20A expression in all three standard scores, including total, cytoplasmic and nuclear, were significantly correlated with FOXM1 expression (P=0.006, P=0.019 and P=0.034, respectively).

183

Figure 4.19: Upregulation of KIF20A significantly associated with poorer survival in breast cancer patients. (A) Kaplan-Meier survival analysis of all patients (n=100) and (B) patients received chemotherapy (n=60), compared by log-rank test, revealed that nuclear overexpression of KIF20A was significantly correlated with poorer survival. (P=0.045 and P=0.016, respectively). (C) The other Kaplan-Meier survival analysis of KIF20A and FOXM1 mRNA transcript expression in a previously published cohort (n= 1,809) (Györffy et al., 2010) also showed that the overexpression of FOXM1 and KIF20A mRNA significantly associated with poorer survival.

184

185

Figure 4.20: Univariate and multivariate Cox-regression analysis using KIF20A nuclear score and other clinicopathological parameters. (A) Nuclear KIF20A overexpression is associated with higher risk of death in univariate analysis (P = 0.05, RR=2.141 for overall survival and P=0.02, RR=2.695 for disease-specific survival, respectively) and remained associated with poor survival after correcting for tumour stage and lymph-node involvement (P = 0.047, RR=2.47 for overall survival and P=0.037, RR=2.767 for disease-specific survival, respectively) (B) Nuclear KIF20A overexpression significantly associated with poorer survival in breast cancer patients who had received chemotherapy (log-rank test, P = 0.008 for overall survival and P=0.004 for disease-specific survival, respectively). In this cohort, Cox-regression analysis showed that nuclear KIF20A overexpression is an even stronger risk marker (P=0.013, RR=4.008 for overall survival and P=0.01, RR=5.089 for disease-specific survival, respectively).

186

4.3 Discussion

Mitosis is a complicated process in actively proliferating cells, leading to the division into duplicated sets of chromosomes and two genetically identical daughter cells.

Failure of cell cycle regulation often causes genetic instability, culminating either in cell death or in tumorigenesis (Chan et al., 2012). As the hallmark of cancer is involved in cell cycle deregulation, anti-mitotic therapies therefore are effective against the abnormal proliferation of transformed cells (Chan et al., 2012). Paclitaxel is a powerful anti-mitotic agent which targets the microtubule dynamic, leading to mitotic arrest. Cells arrested in mitosis either die during mitosis or undergo a process called mitotic slippage, where the cells enter G1 without undergoing anaphase or cytokinesis to produce a single, tetraploid cell. Despite the success of paclitaxel, toxic side effects and the development of drug resistance represent main obstacles to improve the overall response and survival of breast cancer patients (Orr et al.,

2003). To overcome this problem, it is thus necessary to better understand the mechanisms of drug action and resistance.

In this chapter, I found that FOXM1 is overexpressed in paclitaxel-resistant MCF-7

TaxR breast cancer cell lines when compared with the parental sensitive MCF-7 cells. FOXM1 expression is downregulated in response to paclitaxel in MCF-7 cells, but remains persistently high in the resistant cells following paclitaxel treatment

(Figure 4.4). These data suggests the possibility that FOXM1 is a target of paclitaxel and that it has a role in mediating paclitaxel action and resistance. Consistent with this notion, the previous result from our laboratory has shown that paclitaxel exerts its cytotoxicity through FOXO3a, which is an upstream negative regulator of FOXM1 expression and activity (Karadedou et al., 2012, Sunters et al., 2006).

187

It is relatively well established that the elimination of cancer cells by high concentrations of paclitaxel is predominantly mediated by triggering apoptosis

(Chang et al., 1999, Roninson et al., 2001, Blagosklonny and Fojo, 1999, Zasadil et al., 2014). Interestingly, this study reveal that in response to low concentrations of paclitaxel (≤10 nM), breast cancer cell lines undergo mitotic catastrophe (aberrant mitosis) which have typical features including mitotic spindle disorganisation, failed chromosome segregation, and formation of multinucleated cells. This aberrant mitotic process is followed by cellular senescence and/or non-apoptotic cell death.

Crucially, similar to paclitaxel treatment, FOXM1 depletion also induces mitotic catastrophe, culminating in non-apoptotic cell death and senescence (Figures 4.10 and 4.12). Consistently, a previous study also showed that silencing of FOXM1 can induce centrosome amplification and mitotic catastrophe as well as senescence in cancer cells (Laoukili et al., 2005, Wonsey and Follettie, 2005, Zhao et al., 2014).

During mitosis, the bipolar mitotic spindle assembly is essential for ensuring appropriated chromosome segregation and, hence, for successful cell division (Chee and Haase, 2010). Kinesin proteins are the main factors required for proper spindle formation, chromosome alignment and segregation (Chee and Haase, 2010).

Kinesin proteins have been proposed to control microtubule dynamics within a mitotic spindle (Walczak and Mitchison, 1996). KIF20A (Kinesin-like family member

20A), also known as MKPL2 or Rab6 kinesin, belongs to the family of kinesin proteins which was originally involved in Golgi apparatus dynamics through direct interaction with Rab6 small GTPase. Recent reports have shown that KIF20A accumulates in mitotic cells and is thought to be essential for cell cycle regulation during cytokinesis (Zhang et al., 2014b). Here, I found that KIF20A is transcriptionally regulated by FOXM1. In agreement, knockdown of FOXM1 in MCF-7 breast cancer

188 cell lines as well as in MEFs led to a significant downregulation of KIF20A at both protein and mRNA levels, whereas overexpression of FOXM1 enhanced KIF20A expression level (Figure 4.5). Moreover, Chromation-Immunoprecipitation (ChIP) and luciferase assay results showed that FOXM1 also specifically binds and regulates the transcriptional activity of KIF20A gene (Figure 4.7- 4.8).

In MCF-7 cells, the expression of KIF20A, like FOXM1, is simultaneously downregulated in response to paclitaxel treatment, suggesting the possibility that paclitaxel may exert its cytotoxic action by suppressing the expression of both

FOXM1 and KIF20A expression. Consistent with this idea, I found that KIF20A or

FOXM1 knockdown using siRNA results in an increase in the frequency of abnormal mitotic spindle formation with non-aligned or misaligned chromosomes (Figure 4.10).

These results are consistent with the previous study performed in human hepatoma cell lines, showing that KIF20A depleted cells exhibited abnormal mitosis, with misaligned chromosomes in metaphase and reduced furrow ingression in telophase and cytokinesis (Gasnereau et al., 2012). Moreover, studies in Drosophila also showed that mutants of Subito, an ortholog of KIF20A in mammalian cells, improperly assembled microtubules at metaphase, leading to activation of the spindle assembly checkpoint and lagging chromosomes at anaphase. This further suggests that KIF20A participates in mitotic spindle assembly and interacts with mitotic regulators (Cesario et al., 2006, Zhang et al., 2014b). However, the paclitaxel-induced spindle abnormalities and chromosome misalignment defects are not further enhanced by depletion of KIF20A or FOXM1, further confirming the idea that paclitaxel targets the FOXM1-KIF20A axis to induce mitotic catastrophe in MCF-

7 cells (Figure 4.10). Interestingly, our data also showed an increase in mitotic catastrophe which occurred spontaneously in some FOXM1 or KIF20A depleted

189 cells, resembling paclitaxel treatment. This further supports that paclitaxel exerts its function through suppressing FOXM1 and KIF20A expression. The spindle assembly checkpoint (SAC) ensures proper chromosome segregation and delays the transition to anaphase in the presence of is defective mitotic spindle formation with unaligned chromosomes (Wolanin et al., 2010, Musacchio and Salmon, 2007). When the SAC functions properly, prolonged incubation with paclitaxel causes abnormal mitotic spindle formation and metaphase arrest leading to mitotic catastrophe. Therefore, I speculate that the elevated expression of FOXM1 and KIF20A observed in the paclitaxel resistant cells counteracts the ability of paclitaxel to induce abnormal spindles through their downregulation.

In term of paclitaxel resistance, similar to FOXM1, KIF20A protein was upregulated in paclitaxel resistant breast cancer cell, indicating that both FOXM1 and KIF20A might contribute to paclitaxel resistance. I next tested the efficiency of targeting

FOXM1 and KIF20A in MCF-7 and the paclitaxel-resistant MCF-7 TaxR breast cancer cell lines. The clonogenic assay results showed that FOXM1-knockdown sensitised MCF-7 cells to long-term proliferative arrest at relatively low paclitaxel dose (1 and 3 nM). Like FOXM1 depletion, knockdown of KIF20A also sensitised

MCF-7 cells to paclitaxel at the very low concentrations of the drug (Figure 4.14).

Interestingly, knockdown of FOXM1 or KIF20A alone almost completely abolished the colony-forming capacity of MCF-7 TaxR cells irrespective of the dosage of paclitaxel used, suggesting that MCF-7 TaxR cells are dependent on the high expression levels of FOXM1 and KIF20A for long-term survival (Figure 4.17A).

Consistent with the clonogenic assay results, FOXM1 and KIF20A knockdown in

MCF-7 TaxR cells significantly induced senescence-associated SA-βgal activity and morphology independent of the paclitaxel concentration, suggesting that MCF-7

190

TaxR cells have become dependent on the high expression of FOXM1 and KIF20A to override non-apoptotic cell death (Figure 4.17B). These findings also appear to be consistent with a previous study that identified KIF20A as effective inducer of non- apoptosis cell death (Groth-Pedersen et al., 2012).

Finally, in order to evaluate the potential ability of FOXM1 and KIF20A as prognostic and predictive biomarkers, I validated my in vitro findings in breast cancer patient samples by using immunohistochemistry or Tissue Microarray Analysis (TMA).

Statistical analysis of the expression patterns revealed that there was a strong and significant correlation between the nuclear staining of FOXM1 and total KIF20A staining, providing further physiological evidence that FOXM1 regulates KIF20A expression in breast cancer patient samples. Crucially, like FOXM1, nuclear KIF20A overexpression (Figure 4.18) is also correlated with poor prognosis in terms of overall and disease specific survival. As 60% of these patients studied have received chemotherapy in the forms of anthracyclines and/or taxanes, these immunohistochemistry data also support the idea that FOXM1 regulates KIF20A to modulate paclitaxel resistance. In summary, my data suggest that paclitaxel targets

FOXM1 to downregulate mitotic kinesins, such as KIF20A, to disrupt normal mitotic spindle formation, thus inducing senescence-related cell cycle arrest and cell death in these cells.

191

CHAPTER 5

FINAL DISCUSSION

192

5. Final Discussion

Overexpression of the oncogenic transcription factor FOXM1 has been implicated in the development of drug resistance to various cytotoxic agents including epirubicin

(Millour et al., 2011, Monteiro et al., 2013, Myatt et al., 2014) and paclitaxel (Carr et al., 2010, Li et al., 2014). However, the exact molecular mechanisms underlying the regulation of sensitivity and resistance to these drugs by FOXM1 is not completely understood.

At the molecular level, the oncogenic potential of FOXM1 is mainly associated with its ability to transcriptionally regulate genes which are involved in different aspects of cancer development (Halasi and Gartel, 2013a, Koo et al., 2012). Therefore, in order to have a better understanding of how FOXM1 contributes to drug resistance, it is imperative to identify novel downstream targets, which would be differently expressed in sensitive and resistance breast cancer cells. The identification of novel targets of FOXM1 will not only be crucial to provide better insights into the precise mechanism of the action of chemotherapeutic drugs, but would also provide the basis for improving drug design aiming to specifically eradicate cancer cells with a limited toxicity towards normal cells. In this thesis, I have identified and characterised two novel FOXM1 downstream targets - NBS1 and KIF20A that are involved in epirubicin and paclitaxel drug resistance, respectively.

In chapter 3, I have established that FOXM1 mediates acquired epirubicin resistance in breast cancer cells through regulating NBS1, a DNA repair gene. This enhances

HR repair activity to alleviate epirubicin-induced DNA damage (Khongkow et al.,

2014). Since the mechanism of action of epirubicin is associated with the induction of

DNA double-strand breaks (DSBs) (Minotti et al., 2004, Nardella et al., 2011, Lord

193 and Ashworth, 2012, Choudhury et al., 2009), I investigated the involvement of

FOXM1 in DNA damage response signalling and found that FOXM1 directly regulates NBS1 expression through a Forkhead response element (FHRE) in its promoter and in turn, upregulation of NBS1 could enhance the stability of MRN complex, leading to an increased kinase activity of ATM. Together with the results of

DR-GFP repair assay, this clearly confirms that the functional role of FOXM1 in HR- mediated DSB repair is directly linked to its ability to regulate NBS1 expression.

Additionally, FOXM1 has previously been reported to indirectly enhance HR repair ability through promoting Skp2 and Cks1 (Wang et al., 2005a), which are key components of the Skp2–SCF E3 ligase complex that mediates the K63-linked ubiquitination of NBS1 upon DSBs. This process is essential for the interaction of

NBS1 with ATM and leads to the activation of ATM and its recruitment to the DNA damage sites to initiate HR repair (Zona et al., 2014, Wu et al., 2012). Furthermore,

FOXM1 has also been reported to regulate several other DNA repair effector proteins, such as BRIP1 (Monteiro et al., 2013), RAD51 (Zhang et al., 2012), EXO1

(Zhou et al., 2014), BRCA2, and XRCC1 (Tan et al., 2007). Besides regulating the expression of DNA damage sensor and effector proteins, FOXM1 has also been shown to modulate chromatin structure remodelling and DNA repair by regulating the expression of SIRT1 (Zona et al., 2014). This protein functions to deacetylate NBS1, enabling it to activate ATM signalling (Yuan et al., 2007). Altogether, these findings strongly suggest that FOXM1 plays a critical role in DNA damage response through regulating NBS1 expression and ATM activity. This function of FOXM1 as transcription regulator of DNA repair genes makes FOXM1 an attractive target to enhance the cellular sensitivity of tumour cells to chemotherapy or radiotherapy- induced DNA damage.

194

In agreement with this finding, I also found that depletion of FOXM1 in both MEFs and MCF-7 cells caused a defect in DNA repair in response to epirubicin, as indicated by the higher level of γH2AX foci. Conversely, reconstitution of FOXM1 in

Foxm1-deficient MEFs decreased the accumulation of γH2AX foci in response to epirubicin treatment, further supporting that FOXM1 plays a critical role in DNA damage response and protects the cells from drug-induced DNA damage.

Additionally, I found that the protein expression patterns of NBS1 expression was in accordance to FOXM1 in epirubicin sensitive and resistant MCF-7 cells. NBS1 expression was downregulated in MCF-7 cells but persistently upregulated in the resistant cells following epirubicin treatment, suggesting that both FOXM1 and NBS1 play a role in mediating epirubicin action and resistance. In MCF-7, depletion of

NBS1 significantly inhibited long-term cell proliferation and triggered premature senescence in response to low doses of epirubicin. This occurs due to the defects in

DNA repair, leading to accumulation of irreparable DNA lesions, persistent DDR signalling, prolonged p53-dependent cell cycle arrest, and eventually an essentially irreversible senescence arrest (Rodier et al., 2009a). More interestingly, FOXM1 or

NBS1 depletion alone in MCF-7 EpiR cells severely impaired their long-term viability and significantly induced cellular senescence independent of epirubicin concentrations used. This finding suggests that the epirubicin-resistant phenotype in

MCF-7 cells depends on the overexpression of FOXM1 and NBS1 to override epirubicin-induced DNA damage and senescence and thus promote long-term survival (Khongkow et al., 2014). Consistent with this, siRNA targeting NBS1 has been shown to increase radiation sensitivity of human non-small cell lung cancer cells (Ohnishi et al., 2006). NBS1 has previously been shown to play a role in persistent DNA damage-induced senescence-associated with inflammatory cytokine

195 secretion (Rodier et al., 2009a). Interestingly, the expression of an NBS1 mutant, which interrupts the MRN function for DNA repair and decreases HR repair activity, has been reported to be associated with an increased sensitivity to DNA damaging cytotoxic agents in head and neck cancer (Tran et al., 2004, Araki et al., 2010) and leukaemia cells (Rink et al., 2007). In the line with these findings, my results strongly suggest that FOXM1 and its novel target NBS1 can protect MCF-7 EpiR cells from entering senescence/cell death by enhancing DNA repair, making them attractive molecular targets for overcoming DNA damaging cytotoxic agents and radiation resistance (Figure 5.1). However, concentrations used in γH2AX foci assays (Figure

3.1, 3.2 and 3.4) were still high. Further γH2AX staining experiments in MEFs and

MCF-7 cells treated with the lower doses of epirubicin will be required to definitely confirm that low doses of epirubicin together with FOXM1 or NBS1 depletion are sufficient to induce DNA damage and eventually senescence.

In response to DSBs, ATM is the main factor, which facilitates checkpoint activation and DNA repair. ATM phosphorylates NBS1, a member of the MRN complex that reinforces the DDR process and facilitates in DNA repair; CHK2, which promotes growth arrest; and p53, a tumour suppressor and transcriptional regulator that orchestrates DNA repair and cell cycle arrest. DSBs that cannot be repaired cause constitutive DDR signalling, prolonged p53-dependent growth arrest, and eventually an essentially irreversible senescence arrest (Rodier et al., 2009b, Campisi and d'Adda di Fagagna, 2007).

As ATM and the MRN complex play critical roles in mediating the DNA damage response and cell cycle checkpoints, these proteins have been considered as promising targets for radiosensitisation and chemosensitisation (Hosoya and

Miyagawa, 2014). For example, the use of NBS1 C-terminal small peptides, wild-

196 type NBS1 inhibitory peptides (wtNIP), has been demonstrated to specifically inhibit

ATM activation-mediated DNA damage responses and enhance the radiosensitivity of cancer (Cariveau et al., 2007). ‘Mirin’, a novel small-molecule inhibitor of the

MRN-ATM signalling pathway has also been isolated and developed to be used as cellular radio- and chemosensitisation compounds. This inhibitor prevents ATM activation in response to DSBs and blocks HR-repair in mammalian cells (Dupre et al., 2008). In addition, several promising ATM inhibitors, such as caffeine (Blasina et al., 1999), Wortmannin (Sarkaria et al., 1998), KU55933 and KU559403, were also shown to sensitize cancer cells to a variety of DNA-damaging agents in preclinical studies (Weber and Ryan, 2015). Therefore, targeting FOXM1 together with DNA damage repair inhibitors might be a promising strategy for overcoming DNA damaging cytotoxic agents and radiation resistance.

In the context of paclitaxel sensitivity and resistance, in chapter 4, I identified and characterised KIF20A as a novel downstream target of FOXM, involved in mitotic spindle assembly and paclitaxel resistance (Khongkow et al., 2015). Firstly, I showed that FOXM1 is overexpressed in paclitaxel-resistant MCF-7 cells and its depletion sensitised both sensitive and resistant cells to paclitaxel by promoting cellular senescence. Additionally, I also found that in MCF-7 cells, the expression of FOXM1 and KIF20A was simultaneously downregulated in response to paclitaxel treatment, suggesting that paclitaxel may exert its cytotoxic action by suppressing the expression of these proteins. Since paclitaxel functions primarily by interfering with spindle microtubule dynamics, causing cell cycle arrest and apoptosis (Montero et al., 2005), I also investigated the functional effects of silencing both FOXM1 and

KIF20A. Depletion of KIF20A or FOXM1 caused an increase in the frequency of abnormal mitotic spindle formation with non-aligned or misaligned chromosomes. In

197 the line with this finding, a specific small molecule inhibitor of KIF20A called paprotrain has also been shown to induce the accumulation of the cells at metaphase and anaphase and exhibit an increased percentage of misaligned chromosomes as well as multipolar spindles (Tcherniuk et al., 2010). However, the paclitaxel-induced spindle abnormalities and chromosome misalignment defects were not increased by depletion of KIF20A or FOXM1, further confirming that paclitaxel targets the FOXM1-KIF20A axis to induce mitotic catastrophe in MCF-7 cells. The spindle assembly checkpoint (SAC) ensures proper chromosome segregation and delays the transition to anaphase in the presence of defective mitotic spindle formation with unaligned chromosomes (Wolanin et al., 2010,

Musacchio and Salmon, 2007). When SAC functions properly, prolonged treatment with paclitaxel leads to abnormal mitotic spindle formation and metaphase arrest, resulting in mitotic catastrophe. Similar to FOXM1, KIF20A protein was upregulated in MCF-7 TaxR cells. Therefore, I speculated that the upregulation of FOXM1 and

KIF20A observed in the paclitaxel resistant cells could counteract the inhibitory effects of paclitaxel on mitotic spindle formation and proliferation through their downregulation. In agreement with this, I found that the silencing of FOXM1 or

KIF20A alone was sufficient to dramatically impair the colony-forming capacity and induce the SA-βgal activity in MCF-7 TaxR cells independent of the paclitaxel concentration used. This suggested that MCF-7 TaxR cells are dependent on the elevated expression levels of FOXM1 and KIF20A, which are essential for long-term survival and senescence suppression. These findings also appear to be consistent with a previous study that has identified KIF20A as an effective inducer of non- apoptosis cell death (Groth-Pedersen et al., 2012). Furthermore, KIF20A has been shown to be crucial in immortalised human fibroblasts undergoing senescence upon

198 activation of the p16/Rb and p53/p21 pathways (Rovillain et al., 2011, Whitfield et al.,

2002).

Therefore, targeting FOXM1 together with Kinesin inhibitors, such as paprotrain

(Tcherniuk et al., 2010) or Eg5 inhibitors (Exertier et al., 2013) could be a promising strategy for treating paclitaxel resistant cells. However, the exact functional link between FOXM1 and KIF20A in mitotic spindle assembly remains elusive and requires further investigation. In addition, KIF2C (Zhao et al., 2014, Ganguly et al.,

2011), PLK1 (Hou et al., 2013, Feng et al., 2012), and Stathmin (Carr et al., 2010,

Alli et al., 2007) have also been identified as downstream regulators of FOXM1, participated in mitotic control and paclitaxel resistance.

More importantly, the clinical significance of the regulation of NBS1 by FOXM1 was further validated by using tissue microarrays (TMAs). The significant correlation between nuclear FOXM1 and total NBS1 expression was observed in breast cancer patient samples. High expression levels of FOXM1 and NBS1 were significantly correlated with poor prognosis in breast cancer, supporting a physiological role of

FOXM1 and NBS1 in genotoxic drug resistance (Khongkow et al., 2014). Consistent with this, NBS1 overexpression has previously been identified as a prognostic marker of some human cancers, including head and neck squamous cell carcinoma

(HNSCC) (Yang et al., 2006, Yang et al., 2005), oral squamous cell carcinoma

(OSCC) (Hsu et al., 2010), and prostate cancer (Cybulski et al., 2013).

Together, the findings in chapter 3 led me to conclude that NBS1 is an important novel target of FOXM1 involved in HR-mediated DSB repair, DNA damage-induced senescence and epirubicin resistance (Figure 5.1). These findings also provide the potential therapeutic benefit of FOXM1-NBS1 axis as a promising target for

199 therapeutic intervention to overcome epirubicin resistance and a reliable prognostic marker used to monitor treatment efficiency in breast cancer.

Using the same set of patient samples, similar to FOXM1, the elevated expression of nuclear KIF20A was associated with poor prognosis in terms of overall and disease specific survival in these breast cancer patient samples. As 60% of these patients have received chemotherapy in the forms of either anthracyclines or taxanes, these immunohistochemistry data also supports the notion that FOXM1 regulates KIF20A to modulate paclitaxel resistance (Khongkow et al., 2015). Accordingly, my findings in chapter 4 suggest that paclitaxel targets FOXM1 to downregulate KIF20A to disrupt the formation of mitotic spindle, thus inducing senescence and cell death in breast cancer cells (Figure 5.2).

Importantly, this work supports that FOXM1, NBS1 and KIF20A could be useful biomarkers for predicting and monitoring chemotherapy response. Through the depletion of FOXM1 and its targets, it is possible that chemotherapy resistance can be reversed, and these proteins could be new therapeutic targets in chemotherapy- resistant breast cancer.

Many advances have been achieved by the use of chemotherapeutic drugs in the treatment of breast cancer patients and in many cases these advances have resulted in improved survival. Despite initial responses, cancer cells acquire resistance to cytotoxic and cytostatic agents reducing the efficacy of current interventions and as a result patients relapse (Chabner and Roberts, 2005).

Traditionally, cancer therapy depends on the cytotoxic effects of chemotherapeutic drugs that aim at causing cell death within a tumour but have severe toxic side effects in patients. In recent years, cellular senescence has been given much

200 attention as a tumour suppressive mechanism. Hence, an alternative treatment option for cancer patients would be to induce cytostasis, disrupting the proliferative capacity of cancer cells without inducing cancer cell death. In line with this, in this thesis I also studied the role of FOXM1 in inducing cellular senescence in the presence of cytostatic concentrations of epirubicin and paclitaxel.

Interestingly, I observed that the concentrations of DNA damaging agents required for inducing senescent phenotypes in MCF-7 as well as in MEF cells are substantially lower than those required to induce cell death (Millour et al., 2011,

Monteiro et al., 2013), further suggesting that cellular senescence may be the predominant mechanism of action for genotoxic anti-cancer agents. Consistent with this, Xue and colleagues has previously revealed that reactivation of endogenous p53 in p53-deficient tumours results in complete tumour regression (Xue et al.,

2007). However, in this case the primary response triggered by p53 was not an apoptotic response, but instead the induction of cellular senescence and these senescent cells can be cleared in vivo through the innate immune response (Xue et al., 2007).

Another interesting point obtained from this study is the fact that the inhibition of

FOXM1 expression sensitises MCF-7 as well as MEF cells into entering epirubicin- induced senescence, as indicated by cellular markers of senescence, including the prolonged accumulation of γH2AX foci, the loss in long-term cell proliferation, the induction of β-galactosidase activity as well as the flat cell morphology. Consistent with my findings, Wang and colleagues have previously reported that early passage

Foxm1-/- MEFs failed to proliferate, displayed a block in mitotic progression and induced premature senescence, as indicated by high expression levels of SA-β gal, p19ARF, p16INK4A, p21CIP1 and p27KIP1(Wang et al., 2005a, Wang et al., 2008, Tan et

201 al., 2007). Moreover, FOXM1 has been identified to be one of the most highly downregulated transcription factor in immortalised human fibroblasts undergoing senescence upon activation of the p16/Rb and p53/p21 pathways and ectopic expression of its constitutive active form (ΔNΔKEN) was sufficient to bypass senescence arrest (Rovillain et al., 2011). Furthermore, depletion of FOXM1 in gastric cancer cells resulted in increased cellular senescence, which was partially dependent on p27KIP1 (Zeng et al., 2009) whereas, CDK4/6 phosphorylation of

FOXM1 has been shown to suppress senescence in melanoma cells (Anders et al.,

2011).

In resistant cells, I observed that FOXM1 depletion alone severely impairs their long- term viability and significantly induced cellular senescence independent of genotoxic drug concentrations used. This finding suggests that the resistant phenotype in

MCF-7 cells depends on the overexpression of FOXM1 to override drug-induced

DNA damage and senescence and thus promote long-term survival. According to the senescence suppression activity of FOXM1, combined targeting of FOXM1 with low doses of chemotherapy could represent a promising strategy to improve patient survival with fewer toxic side effects.

Since elevated expression of FOXM1 confers resistance to various chemotherapeutic agents, targeting FOXM1 is a promising therapeutic intervention to override resistance to genotoxic cancer agents, including epirubicin, paclitaxel, cisplatin and ionising radiation. In recent years, several FOXM1 inhibitors have been identified (Halasi and Gartel, 2013b). For example, the thiazole antibiotics Siomycin

A and thiostrepton have been demonstrated to suppress the transcriptional activity and the expression of FOXM1 via proteasome inhibition and induce apoptosis in human cancer cells (Radhakrishnan et al., 2006, Bhat et al., 2009b). Accumulating

202 evidence have suggested that treatment with Siomycin A and/or thiostrepton caused a significant inhibition of tumour growth in a wide range of human cancer cells including breast (Kwok et al., 2008), melanoma (Bhat et al., 2008), leukaemia, osteosarcoma (Bhat et al., 2009b), brain (Priller et al., 2011), and prostate cells

(Pandit and Gartel, 2010). Other well-known proteasome inhibitors, such as bortezomib, MG132 and MG115 have also demonstrated to inhibit FOXM1 transcriptional activity and expression (Bhat et al., 2009a). Recently, an alternative mechanism of action of thiostrepton has been proposed that thiostrepton directly interacts with FOXM1, blocking its binding to the promoters of its target genes,

(Hegde et al., 2011, Kwok et al., 2008). Furthermore, the synthetic ARF peptide has been shown to effectively suppress FOXM1 transcriptional activity, inhibit FOXM1- induced colony growth and tumour metastasis and also induce apoptosis of several distinct human cancers, including osteosarcoma and hepatocellular carcinoma cells, as well as in mouse models (Kalinichenko et al., 2004, Costa et al., 2005b, Gusarova et al., 2007, Park et al., 2011). Collectively, it seems that FOXM1 could be targeted successfully as a single entity with promising results in the fight against cancer.

However, accumulating data have suggested that combination therapy may improve treatment efficacy as well as lower the drug dose needed for anti-tumour growth effects, thus reducing the adverse side effects and the probability of cancer cells becoming drug-resistant. For example, Kwok and colleagues have demonstrated that thiostrepton synergised with cisplatin to reverse acquired cisplatin resistance in breast cancer cells and caused a substantial induction of cisplatin-induced cell death

(Kwok et al., 2010b). This is the first study suggesting that the use of FOXM1 inhibitor, thiostrepton, in combination with chemotherapy could provide a novel mechanism to reverse the phenomenon of chemoresistance in breast cancer.

203

Another interesting point obtained from this study is the fact that thiostrepton selectively targets breast cancer cells but not non-transformed cells, suggesting that thiostrepton could be a lead compound for targeted therapy of breast cancer with minimal toxicity against non-malignant cells (Kwok et al., 2008, Kwok et al., 2010b).

In addition to this, Park and colleagues have also demonstrated that silencing of

FOXM1 also led to higher sensitivity to doxorubicin in breast cancer cells as well as in a xenograft mouse model. They also suggested that FOXM1-dependent resistance to doxorubicin is mediated by regulating DNA repair genes and proposed that FOXM1 is a poor prognosis transcription factor conferring doxorubicin resistance in breast cancer (Park et al., 2012). Furthermore, recent studies from our lab have also revealed that FOXM1 was found to be overexpressed in epirubicin resistance cells and its depletion was able to resensitise these cells to epirubicin (Millour et al.,

2011, Monteiro et al., 2013, Khongkow et al., 2014). We also found that FOXM1- dependent epirubicin resistance was partially mediated by enhanced expression of

DNA repair genes, including BRIP1 (Monteiro et al., 2013). Additionally, my findings in this thesis also demonstrate that NBS1 and KIF20A are crucial targets of FOXM1 in mediating resistance to epirubicin and paclitaxel, respectively (Khongkow et al.,

2014, Khongkow et al., 2015). In accordance, an independent study by Carr and colleagues reported that FOXM1 overexpression confers resistance to herceptin and paclitaxel resistance cells and its depletion by the ARF–derived peptide inhibitor was able to resensitise FOXM1 overexpressing breast cancer cell lines to herceptin and paclitaxel (Carr et al., 2010).

The better understanding of the cellular and molecular mechanisms whereby chemotherapeutic drugs achieve their toxicity effects may allow for the development of better therapeutic choices for patients that have become resistant to treatment. My

204 results indicate that FOXM1 is a crucial cellular target of epirubicin and paclitaxel in breast cancer cells. In accordance, the expression and the activity of FOXM1 was substantially repressed in response to both epirubicin and paclitaxel treatments.

However, the exact molecular mechanism by which they modulate FOXM1 remains mostly elusive. Therefore, it is crucial not only to identify its downstream targets, but also the upstream regulators of FOXM1 in response to chemotherapy treatment.

Targeting upstream regulators of FOXM1, might prove more useful to prevent drug induced DNA damage and senescence than targeting FOXM1 downstream targets or FOXM1 itself. Accumulating evidence suggests that this repression of FOXM1 could occur through transcriptional and post-translational mechanisms.

At the transcriptional level, Millour and colleagues have previously reported that ATM and p53 coordinately regulate FOXM1 expression via its upstream transcriptional regulator E2F1 to modulate epirubicin response and resistance in breast cancer. In sensitive cells, epirubicin suppressed FOXM1 expression at transcriptional levels through the activation of p53 and repression of E2F1 activity (Millour et al., 2011). By contrast, they observed that the loss of p53 function and the increased expression and activity of ATM, lead to an induction of FOXM1 through E2F1 in chemoresistant cells. ATM-E2F1 pathway may have important implications for treatment of drug- resistant cancers, particularly those lacking functional p53. In addition, ATM and

FOXM1 inhibitors could also be used in combination with conventional genotoxic therapeutics to enhance the drug efficacy and for overcoming resistance (Millour et al., 2011). In agreement with this, small-molecule activators of p53, nutlin-3 and

RITA, has been shown to retain functional p53 and inhibit FOXM1 mRNA and protein expression, leading to the increased sensitivity to chemotherapeutic agents (Barsotti and Prives, 2009, Michaelis et al., 2012). Mechanistically, this p53-mediated

205 repression of FOXM1 has also been suggested to partially depend on p21CIP1 and pRB (Barsotti and Prives, 2009). In addition to this, the results from our laboratory have previously shown that paclitaxel and gefitinib exert their cytotoxic functions through FOXO3a which is an upstream negative regulator of FOXM1 expression and activity (Karadedou et al., 2012, McGovern et al., 2009).

Recently, some microRNAs (miRNAs), endogenous small non-coding RNAs are also reported to directly target FOXM1 and influence the oncogenic functions of cancer cells. For example, miR-134 and miR-149 has been suggested to block epithelial to mesenchymal transition (EMT) by targeting FOXM1 in non-small cell lung cancer (Li et al., 2012). In addition, miR-320 has been shown to enhance the sensitivity of human colon cancer cells to chemotherapy in vitro by targeting FOXM1 (Wan et al.,

2015). Thus, altered expression of miRNAs in cancer cells could also act as potential tools for cancer diagnosis, prognosis and treatment.

Another possibility for targeting FOXM1 could rely on interfering with its post- translational modifications to disable its activity. Phosphorylation is one of the main post-translational modifications, which modulates FOXM1 expression, activity and localisation (Myatt and Lam, 2007). Previous studies have demonstrated that treatment with DNA-damaging insults, such as γ-irradiation, etoposide and UV, promotes CHK2-induced phosphorylation of FOXM1. Such phosphorylation of

FOXM1 results in its stabilisation leading to the transcriptional activation of downstream DNA repair genes (Tan et al., 2007). In addition, FOXM1 has been identified as a critical target of the CyclinD-CDK4/6 kinases. This phosphorylation leads to its stabilisation and activation, thereby maintaining its expression and suppressing the levels of reactive oxygen species (Anders et al., 2011).

Furthermore, recent studies have shown that SUMOylation is believed to inhibit

206

FOXM1 activity and delays mitotic transition in response to treatment with epirubicin and mitotic inhibitors in MCF-7 breast cancer cells (Myatt et al., 2014).

In conclusion, my study reveals that FOXM1 is a critical mediator of epirubicin and paclitaxel sensitivity and resistance in breast cancer cells. I also define the involvement of FOXM1 in DNA damage repair, mitotic spindle formation and suppression of senescence, which are vital for cancer progression and drug resistance, through the regulation of its novel downstream targets, NBS1 and

KIF20A. In addition, downregulation of FOXM1 and its targets can enhance the sensitivity of resistance breast cancer cells to the cytotoxic agents, thus suggesting that FOXM1, NBS1, and KIF20A could be novel therapeutic targets for overcoming chemotherapy resistance. According to senescence suppression activity of FOXM1, the combination of FOXM1 inhibitors with chemotherapeutic agents or IR may also lead to the use of lower effective doses and reduced side effects for the breast cancer patients. Apart from mediating chemotherapy resistance, FOXM1, NBS1 and

KIF20A could also be used as biomarkers to predict and monitor the response of breast cancer patients to particular chemotherapeutic agents including epirubicin, and paclitaxel. However, additional validation studies are still required to confirm these findings in other breast cancer subtypes, in different cancer cell lines, in in vivo biological models, as well as in the other cohorts of clinical samples. In addition, the relationship between FOXM1, NBS1, and KIF20A should also be examined in parallel to test whether these three proteins are the main factors mediating resistance to both epirubicin and paclitaxel treatments. Moreover, questions about the expression statuses of FOXM1 isoforms and their impact on breast cancer development and drug resistance should be addressed in future research.

Furthermore, future studies should also uncover the molecular details of the post-

207 translational modification of these proteins in the context of chemotherapy sensitivity and resistance. These, together with the development of more effective and specific inhibitors, will definitely assist to open up the excellent therapeutic window in which targeting of FOXM1 or its targets can achieve clinical benefits.

208

Figure 5.1 Schematic representation summarising the possible role of FOXM1 in the development of epirubicin resistance. In epirubicin sensitive MCF-7 cells, the expression of FOXM1 and its downstream targets, including NBS1 are downregulated by epirubicin, leading to the impairment of HR repair and the induction of cellular senescence and cell death. Conversely, FOXM1 is overexpressed in resistant cells, resulting in the enhancement of HR repair through the upregulation of DNA repair genes, including NBS1. This leads to the increased cell survival and drug resistance.

209

Figure 5.2 Schematic representation summarising the possible role of FOXM1 in the development of paclitaxel resistance. In paclitaxel sensitive MCF-7 cells, the expression of FOXM1 and KIF20A are downregulated by paclitaxel, leading to abnormal mitotic spindle formation and the induction of mitotic catastrophe, cellular senescence or cell death. Conversely, FOXM1 and KIF20A are overexpressed in resistant cells. These proteins might counteract paclitaxel-induced abnormal mitotic spindle formation and also suppress cellular senescence leading to the increased cell survival and drug resistance.

210

SUPPLEMENTARY DATA

211

Figure S1: Quantification of γH2AX foci using FociCounter software (Jucha et al., 2010).

212

Figure S2: FOXM1 depletion leads to increased levels of DNA damage in MCF-7. (A) MCF-7 cells were transfected with NS siRNA or with FOXM1 siRNA for 24 h. Cells were cultured on chamber slides and treated with 1 μM of epirubicin for 0, 4, 24, 48 and 72 h and then stained for γH2AX (green) and DAPI (blue). (B) The graph below shows quantification of γH2AX foci number. Bars represent an average of three independent experiments ± S.D. Statistical analyses were conducted using Student’s t-tests against the correspondent time point (***P≤0.0001, significant; n.s., non-significant).

213

Figure S3: Optimization of epirubicin treatment doses used in long-term clonogenic assay for WT and Foxm1−/− MEFs. Bars represent average ± SD from three independent experiments. Statistical significance was determined by two-tailed unpaired Student’s t-test (**P ≤ 0.01, significant; ns, non-significant).

214

Figure S4: Epirubicin resistant cell line exhibits increased FOXM1 and NBS1 protein. Densitometric analysis of the western blots in figure 3.7A was carried out using Image J software. The intensity of the bands of FOXM1, NBS1, P-NBS1, ATM and P-ATM at each time point was normalised to the intensity of the β-Tubulin band at the respective time point and the obtained values were plotted (n=3). Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test test (*P ≤ 0.05, **P≤0.01, ***P≤0.005 significant; ns, non-significant).

215

Figure S5: MCF-7 and MCF-7 EpiR cells depleted in FOXM1 expression revealed significant decreases at protein level of NBS1. Densitometric analysis of the western blots in figure 3.8 was carried out using Image J software. The intensity of the bands of FOXM1, NBS1, P-NBS1, MRE11, RAD50, ATM and P-ATM at each time point was normalised to the intensity of the β-Tubulin band at the respective time point and the obtained values were plotted (n=3). Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test test (*P ≤ 0.05, **P≤0.01, significant; ns, non-significant).

216

Figure S6: FOXM1 and KIF20A expression levels were strongly upregulated in paclitaxel- resistant MCF-7 cells. Densitometric analysis of the western blots in figure 4.4A was carried out using Image J software. The intensity of the bands of FOXM1 and KIF20A at each time point was normalised to the intensity of the β-Tubulin band at the respective time point and the obtained values were plotted (n=3). Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test test (*P ≤ 0.05, **P≤0.01, significant; ns, non-significant).

217

Figure S7: Silencing of FOXM1 downregulates KIF20A protein expression in MCF-7 and MCF-7 TaxR cells. Densitometric analysis of the western blots in figure 4.5B was carried out using Image J software. The intensity of the bands of FOXM1 and KIF20A at each time point was normalised to the intensity of the β-Tubulin band at the respective time point and the obtained values were plotted (n=3). Bars represent average ± SD. Statistical significance was determined by two-tailed unpaired Student’s t-test test (*P ≤ 0.05, **P≤0.01, significant; ns, non-significant).

218

Figure S8: Knockdown of FOXM1 or KIF20A leads to defects in mitotic spindle formation in MCF-7 TaxR cells. MCF-7 TaxR cells were transfected with NS, FOXM1 or KIF20A siRNA. 24 h after transfection, cells cultured on chamber slides were treated with 5 nM paclitaxel for 24 h. Cells were then fixed, permeabilised, and immunostained with antibody against α-tubulin (Green). Mitotic cells were visualised with Leica TCS SP5 (63X magnification). For each condition, images of at least 50 mitotic cells were captured. Representative confocal images are shown. The number of mitotic cells classified into either normal bipolar, abnormal bipolar, monopolar or multipolar spindles was quantified. Results represent average of three independent experiments ± S.D.

219

Figure S9. FOXM1 was unable to transactivate a putative KIF20A promoter region containing the 5’ UTR (-1150 to -61) in the MCF-7 cells. (A) Schematic representation of a putative KIF20A promoter region containing the 5’ UTR located between -1150 to -61. MCF-7 cells were transiently co- transfected the pGL3 basic luciferase reporter plasmid containing the putative KIF20A promoter, the Renilla luciferase plasmid and the FOXM1B expression plasmid. Twenty-four hours after transfection, cells were lysed, and the luciferase activity examined. Firefly luminescence signal was normalized with the Renilla luminescence signal. (B) Bars represent mean±S.D from three independent experiments.

220

Figure S10: Cloning of WT and mut KIF20A luciferase reporter constructs. The KIF20A-WT, KIF20A-MUT1, and KIF20A-MUT2 fragments were initially designed and synthesised by GeneArt (Life Technologies, Darmstadt, Germany). The synthesised DNA fragments cleaved with the restriction enzymes XhoI and Bglll and cloned into the XhoI/Bglll sites of the pGL3-basic vector.

221

Figure S11: Depletion of FOXM1 increases p21 expression and induces G/M arrest in both MCF-7 and MCF-7 TaxR cells. (A) and (B) MCF-7 and MCF-7 TaxR cells depleted in FOXM1 expression revealed significant increase in p21 expression. (C) and (D) Knockdown of FOXM1 induces G2/M arrest in both MCF-7 and MCF-7 TaxR cells.

222

REFERENCES . AHMAD, A., WANG, Z., KONG, D., ALI, S., LI, Y., BANERJEE, S., ALI, R. & SARKAR, F. 2010. FoxM1 down-regulation leads to inhibition of proliferation, migration and invasion of breast cancer cells through the modulation of extra-cellular matrix degrading factors. Breast Cancer Research and Treatment, 122, 337-346. ALI, S. & COOMBES, R. C. 2002. Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer, 2, 101-112. ALLI, E., YANG, J.-M., FORD, J. M. & HAIT, W. N. 2007. Reversal of Stathmin-Mediated Resistance to Paclitaxel and Vinblastine in Human Breast Carcinoma Cells. Molecular Pharmacology, 71, 1233-1240. ALLISON, K. H. 2012. Molecular Pathology of Breast Cancer: What a Pathologist Needs to Know. American Journal of Clinical Pathology, 138, 770-780. ALIMONTI, A., NARDELLA, C., CHEN, Z., CLOHESSY, J. G., CARRACEDO, A., TROTMAN, L. C., CHENG, K., VARMEH, S., KOZMA, S. C., THOMAS, G., ROSIVATZ, E., WOSCHOLSKI, R., COGNETTI, F., SCHER, H. I. & PANDOLFI, P. P. 2010. A novel type of cellular senescence that can be enhanced in mouse models and human tumor xenografts to suppress prostate tumorigenesis. The Journal of Clinical Investigation, 120, 681-693. AMBUDKAR, S. V., DEY, S., HRYCYNA, C. A., RAMACHANDRA, M., PASTAN, I. & GOTTESMAN, M. M. 1999. BIOCHEMICAL, CELLULAR, AND PHARMACOLOGICAL ASPECTS OF THE MULTIDRUG TRANSPORTER1. Annual Review of Pharmacology and Toxicology, 39, 361-398. ANAND, S., PENRHYN-LOWE, S. & VENKITARAMAN, A. R. 2003. AURORA-A amplification overrides the mitotic spindle assembly checkpoint, inducing resistance to Taxol. Cancer Cell, 3, 51-62. ANDERS, L., KE, N., HYDBRING, P., CHOI, YOON J., WIDLUND, HANS R., CHICK, JOEL M., ZHAI, H., VIDAL, M., GYGI, STEPHEN P., BRAUN, P. & SICINSKI, P. 2011. A Systematic Screen for CDK4/6 Substrates Links FOXM1 Phosphorylation to Senescence Suppression in Cancer Cells. Cancer Cell, 20, 620-634. ARAKI, K., YAMASHITA, T., REDDY, N., WANG, H., ABUZEID, W. M., KHAN, K., O'MALLEY, B. W., JR. & LI, D. 2010. Molecular disruption of NBS1 with targeted gene delivery enhances chemosensitisation in head and neck cancer. Br J Cancer, 103, 1822-1830. ARYA, A. K., EL-FERT, A., DEVLING, T., ECCLES, R. M., ASLAM, M. A., RUBBI, C. P., VLATKOVIĆ, N., FENWICK, J., LLOYD, B. H., SIBSON, D. R., JONES, T. M. & BOYD, M. T. 2010. Nutlin-3, the small-molecule inhibitor of MDM2, promotes senescence and radiosensitises laryngeal carcinoma cells harbouring wild-type p53. British Journal of Cancer, 103, 186-195. ASAHARA, S., TAKEDA, K., YAMAO, K., MAGUCHI, H. & YAMAUE, H. 2013. Phase I/II clinical trial using HLA-A24-restricted peptide vaccine derived from KIF20A for patients with advanced pancreatic cancer. Journal of Translational Medicine, 11, 291-291. BAKKENIST, C. J. & KASTAN, M. B. 2003. DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation. Nature, 421, 499-506. BARSOTTI, A. M. & PRIVES, C. 2009. Pro-proliferative FoxM1 is a target of p53-mediated repression. Oncogene, 28, 4295-4305. BARTKOVA, J., TOMMISKA, J., OPLUSTILOVA, L., AALTONEN, K., TAMMINEN, A., HEIKKINEN, T., MISTRIK, M., AITTOMÄKI, K., BLOMQVIST, C., HEIKKILÄ, P., LUKAS, J., NEVANLINNA, H. & BARTEK, J. 2008. Aberrations of the MRE11–RAD50–NBS1 DNA damage sensor complex in human breast cancer: MRE11 as a candidate familial cancer-predisposing gene. Molecular Oncology, 2, 296-316. BEDARD, P. L., DI LEO, A. & PICCART-GEBHART, M. J. 2010. Taxanes: optimizing adjuvant chemotherapy for early-stage breast cancer. Nat Rev Clin Oncol, 7, 22-36. BEN-PORATH, I. & WEINBERG, R. A. 2005. The signals and pathways activating cellular senescence. The International Journal of Biochemistry & Cell Biology, 37, 961-976.

223

BERAL, V., HERMON, C., REEVES, G. & KEY, T. 1996. Breast Cancer Trends in Women in Sweden, the UK, and the USA in Relation to Their Past Use of Oral Contraceptives. In: LI, J. J., LI, S., GUSTAFSSON, J.-Å., NANDI, S. & SEKELY, L. (eds.) Hormonal Carcinogenesis II. Springer New York. BERNSTEIN, K. A. & ROTHSTEIN, R. 2009. At Loose Ends: Resecting a Double-Strand Break. Cell, 137, 807-810. BHAT, U. G., HALASI, M. & GARTEL, A. L. 2009a. FoxM1 Is a General Target for Proteasome Inhibitors. PLoS One, 4, e6593. BHAT, U. G., HALASI, M. & GARTEL, A. L. 2009b. Thiazole Antibiotics Target FoxM1 and Induce Apoptosis in Human Cancer Cells. PLoS One, 4, e5592. BHAT, U. G., ZIPFEL, P. A., TYLER, D. S. & GARTEL, A. L. 2008. Novel anticancer compounds induce apoptosis in melanoma cells. Cell Cycle, 7, 1851-1855. BLAGOSKLONNY, M. V. & FOJO, T. 1999. Molecular effects of paclitaxel: Myths and reality (a critical review). International Journal of Cancer, 83, 151-156. BLASINA, A., PRICE, B. D., TURENNE, G. A. & MCGOWAN, C. H. 1999. Caffeine inhibits the checkpoint kinase ATM. Current Biology, 9, 1135-1138. BORGNE, A. & MEIJER, L. 1996. Sequential Dephosphorylation of p34cdc2 on Thr-14 and Tyr-15 at the Prophase/Metaphase Transition. Journal of Biological Chemistry, 271, 27847-27854. BRAIG, M. & SCHMITT, C. A. 2006. Oncogene-Induced Senescence: Putting the Brakes on Tumor Development. Cancer Research, 66, 2881-2884. BROXTERMAN, H. J., LANKELMA, J. & HOEKMAN, K. 2003. Resistance to cytotoxic and anti- angiogenic anticancer agents: similarities and differences. Drug Resistance Updates, 6, 111- 127. BUGG, B. Y., DANKS, M. K., BECK, W. T. & SUTTLE, D. P. 1991. Expression of a mutant DNA topoisomerase II in CCRF-CEM human leukemic cells selected for resistance to teniposide. Proceedings of the National Academy of Sciences, 88, 7654-7658. BURMA, S., CHEN, B. P., MURPHY, M., KURIMASA, A. & CHEN, D. J. 2001. ATM Phosphorylates Histone H2AX in Response to DNA Double-strand Breaks. Journal of Biological Chemistry, 276, 42462-42467. BUZDAR, A. U. 2003. Advances in Endocrine Treatments for Postmenopausal Women with Metastatic and Early Breast Cancer. The Oncologist, 8, 335-341. CAMPISI, J. 2001. Cellular senescence as a tumor-suppressor mechanism. Trends in Cell Biology, 11, S27-S31. CAMPISI, J. & D'ADDA DI FAGAGNA, F. 2007. Cellular senescence: when bad things happen to good cells. Nat Rev Mol Cell Biol, 8, 729-740. CANMAN, C. E. & LIM, D. S. 1998. The role of ATM in DNA damage responses and cancer. Oncogene, 17, 3301-3308. CARIVEAU, M. J., TANG, X., CUI, X.-L. & XU, B. 2007. Characterization of an NBS1 C-Terminal Peptide That Can Inhibit Ataxia Telangiectasia Mutated (ATM)-Mediated DNA Damage Responses and Enhance Radiosensitivity. Molecular Pharmacology, 72, 320-326. CARNEY, J. P., MASER, R. S., OLIVARES, H., DAVIS, E. M., LE BEAU, M., YATES III, J. R., HAYS, L., MORGAN, W. F. & PETRINI, J. H. J. 1998. The hMre11/hRad50 Protein Complex and Nijmegen Breakage Syndrome: Linkage of Double-Strand Break Repair to the Cellular DNA Damage Response. Cell, 93, 477-486. CARR, J. R., PARK, H. J., WANG, Z., KIEFER, M. M. & RAYCHAUDHURI, P. 2010. FoxM1 Mediates Resistance to Herceptin and Paclitaxel. Cancer Research, 70, 5054-5063. CASTEDO, M., PERFETTINI, J.-L., ROUMIER, T., ANDREAU, K., MEDEMA, R. & KROEMER, G. 2004. Cell death by mitotic catastrophe: a molecular definition. Oncogene, 23, 2825-2837. CESARIO, J. M., JANG, J. K., REDDING, B., SHAH, N., RAHMAN, T. & MCKIM, K. S. 2006. Kinesin 6 family member Subito participates in mitotic spindle assembly and interacts with mitotic regulators. Journal of Cell Science, 119, 4770-4780.

224

CHABNER, B. A. & ROBERTS, T. G. 2005. Chemotherapy and the war on cancer. Nat Rev Cancer, 5, 65- 72. CHAN, C.-H., MORROW, JOHN K., LI, C.-F., GAO, Y., JIN, G., MOTEN, A., STAGG, LOREN J., LADBURY, JOHN E., CAI, Z., XU, D., LOGOTHETIS, CHRISTOPHER J., HUNG, M.-C., ZHANG, S. & LIN, H.-K. 2013. Pharmacological Inactivation of Skp2 SCF Ubiquitin Ligase Restricts Cancer Stem Cell Traits and Cancer Progression. Cell, 154, 556-568. CHAN, D. W., YU, S. Y. M., CHIU, P. M., YAO, K. M., LIU, V. W. S., CHEUNG, A. N. Y. & NGAN, H. Y. S. 2008. Over-expression of FOXM1 transcription factor is associated with cervical cancer progression and pathogenesis. The Journal of Pathology, 215, 245-252. CHAN, K. S., KOH, C. G. & LI, H. Y. 2012. Mitosis-targeted anti-cancer therapies: where they stand. Cell Death Dis, 3, e411. CHANG, B.-D., BROUDE, E. V., DOKMANOVIC, M., ZHU, H., RUTH, A., XUAN, Y., KANDEL, E. S., LAUSCH, E., CHRISTOV, K. & RONINSON, I. B. 1999. A Senescence-like Phenotype Distinguishes Tumor Cells That Undergo Terminal Proliferation Arrest after Exposure to Anticancer Agents. Cancer Research, 59, 3761-3767. CHANG, M. 2012. Tamoxifen Resistance in Breast Cancer. Biomolecules & Therapeutics, 20, 256-267. CHEE, M. K. & HAASE, S. B. 2010. B-Cyclin/CDKs Regulate Mitotic Spindle Assembly by Phosphorylating Kinesins-5 in Budding Yeast. PLoS Genetics, 6, e1000935. CHEN, C. H., CHIEN, C. Y., HUANG, C. C., HWANG, C. F., CHUANG, H. C., FANG, F. M., HUANG, H. Y., CHEN, C. M., LIU, H. L. & HUANG, C. Y. 2009. Expression of FLJ10540 is correlated with aggressiveness of oral cavity squamous cell carcinoma by stimulating cell migration and invasion through increased FOXM1 and MMP-2 activity. Oncogene, 28, 2723-2737. CHEN, H., YANG, C., YU, L., XIE, L., HU, J., ZENG, L. & TAN, Y. 2012. Adenovirus-mediated RNA interference targeting FOXM1 transcription factor suppresses cell proliferation and tumor growth of nasopharyngeal carcinoma. The Journal of Gene Medicine, 14, 231-240. CHEN, J., GOMES, A. R., MONTEIRO, L. J., WONG, S. Y., WU, L. H., NG, T.-T., KARADEDOU, C. T., MILLOUR, J., IP, Y.-C., CHEUNG, Y. N., SUNTERS, A., CHAN, K. Y. K., LAM, E. W. F. & KHOO, U.- S. 2010. Constitutively Nuclear FOXO3a Localization Predicts Poor Survival and Promotes Akt Phosphorylation in Breast Cancer. PLoS One, 5, e12293. CHIEN, A. J. & MOASSER, M. M. 2008. Cellular Mechanisms of Resistance to Anthracyclines and Taxanes in Cancer: Intrinsic and Acquired. Seminars in Oncology, 35, Supplement 2, S1-S14. CHILDS, B. G., BAKER, D. J., KIRKLAND, J. L., CAMPISI, J. & VAN DEURSEN, J. M. 2014. Senescence and apoptosis: dueling or complementary cell fates? CHOUDHURY, A., ZHAO, H., JALALI, F., AL RASHID, S., RAN, J., SUPIOT, S., KILTIE, A. E. & BRISTOW, R. G. 2009. Targeting homologous recombination using imatinib results in enhanced tumor cell chemosensitivity and radiosensitivity. Molecular Cancer Therapeutics, 8, 203-213. CLINGEN, P. H., WU, J. Y. H., MILLER, J., MISTRY, N., CHIN, F., WYNNE, P., PRISE, K. M. & HARTLEY, J. A. 2008. Histone H2AX phosphorylation as a molecular pharmacological marker for DNA interstrand crosslink cancer chemotherapy. Biochemical Pharmacology, 76, 19-27. COGSWELL, J. P., BROWN, C. E., BISI, J. E. & NEILL, S. D. 2000. Dominant-Negative Polo-like Kinase 1 Induces Mitotic Catastrophe Independent of cdc25C Function. Cell Growth Differ, 11, 615- 623. COLLADO, M., BLASCO, M. A. & SERRANO, M. 2007. Cellular Senescence in Cancer and Aging. Cell, 130, 223-233. COSTA, R. H., KALINICHENKO, V. V., HOLTERMAN, A.-X. L. & WANG, X. 2003. Transcription factors in liver development, differentiation, and regeneration. Hepatology, 38, 1331-1347. COSTA, R. H. 2005. FoxM1 dances with mitosis. Nat Cell Biol, 7, 108-110. COSTA, R. H., KALINICHENKO, V. V., MAJOR, M. L. & RAYCHAUDHURI, P. 2005a. New and unexpected: forkhead meets ARF. Current Opinion in Genetics & Development, 15, 42-48. COSTA, R. H., KALINICHENKO, V. V., MAJOR, M. L. & RAYCHAUDHURI, P. 2005b. New and unexpected: Forkhead meets ARF. Current Opinion in Genetics and Development, 15, 42-48.

225

CRUK, http://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer- type/breast-cancer, Accessed 2015. CYBULSKI, C., WOKOŁORCZYK, D., KLUŹNIAK, W., JAKUBOWSKA, A., GÓRSKI, B., GRONWALD, J., HUZARSKI, T., KASHYAP, A., BYRSKI, T., DĘBNIAK, T., GOŁĄB, A., GLINIEWICZ, B., SIKORSKI, A., ŚWITAŁA, J., BORKOWSKI, T., BORKOWSKI, A., ANTCZAK, A., WOJNAR, Ł., PRZYBYŁA, J., SOSNOWSKI, M., MAŁKIEWICZ, B., ZDROJOWY, R., SIKORSKA-RADEK, P., MATYCH, J., WILKOSZ, J., RÓŻAŃSKI, W., KIŚ, J., BAR, K., BRYNIARSKI, P., PARADYSZ, A., JERSAK, K., NIEMIROWICZ, J., SŁUPSKI, P., JARZEMSKI, P., SKRZYPCZYK, M., DOBRUCH, J., DOMAGAŁA, P., NAROD, S. A., LUBIŃSKI, J. & THE POLISH HEREDITARY PROSTATE CANCER, C. 2013. An inherited NBN mutation is associated with poor prognosis prostate cancer. British Journal of Cancer, 108, 461-468. D'ADDA DI FAGAGNA, F., REAPER, P. M., CLAY-FARRACE, L., FIEGLER, H., CARR, P., VON ZGLINICKI, T., SARETZKI, G., CARTER, N. P. & JACKSON, S. P. 2003. A DNA damage checkpoint response in telomere-initiated senescence. Nature, 426, 194-198. DA SILVA, L. & LAKHANI, S. R. 2010. Pathology of hereditary breast cancer. Mod Pathol, 23, S46-S51. DAI, B., KANG, S. H., GONG, W., LIU, M., ALDAPE, K. D., SAWAYA, R. & HUANG, S. 2007. Aberrant FoxM1B expression increases matrix metalloproteinase-2 transcription and enhances the invasion of glioma cells. Oncogene, 26, 6212-6219. DARZYNKIEWICZ, Z., TRAGANOS, F. & WLODKOWIC, D. 2009. Impaired DNA Damage Response – an Achilles’ Heel Sensitizing Cancer to Chemotherapy and Radiotherapy. European journal of pharmacology, 625, 143-150. DE BRUIN, E. C. & MEDEMA, J. P. 2008. Apoptosis and non-apoptotic deaths in cancer development and treatment response. Cancer Treatment Reviews, 34, 737-749. DI LEONARDO, A., LINKE, S. P., CLARKIN, K. & WAHL, G. M. 1994. DNA damage triggers a prolonged p53-dependent G1 arrest and long-term induction of Cip1 in normal human fibroblasts. Genes & Development, 8, 2540-2551. DIMRI, G. P., LEE, X., BASILE, G., ACOSTA, M., SCOTT, G., ROSKELLEY, C., MEDRANO, E. E., LINSKENS, M., RUBELJ, I. & PEREIRA-SMITH, O. 1995a. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proceedings of the National Academy of Sciences, 92, 9363-9367. DIMRI, G. P., LEE, X., BASILE, G., ACOSTA, M., SCOTT, G., ROSKELLEY, C., MEDRANO, E. E., LINSKENS, M., RUBELJ, I., PEREIRA-SMITH, O., PEACOCKE, M. & CAMPISI, J. 1995b. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proceedings of the National Academy of Sciences of the United States of America, 92, 9363-9367. ECHARD, A., JOLLIVET, F., MARTINEZ, O., LACAPÈRE, J.-J., ROUSSELET, A., JANOUEIX-LEROSEY, I. & GOUD, B. 1998. Interaction of a Golgi-Associated Kinesin-Like Protein with Rab6. Science, 279, 580-585. EFEYAN, A., ORTEGA-MOLINA, A., VELASCO-MIGUEL, S., HERRANZ, D., VASSILEV, L. T. & SERRANO, M. 2007. Induction of p53-Dependent Senescence by the MDM2 Antagonist Nutlin-3a in Mouse Cells of Fibroblast Origin. Cancer Research, 67, 7350-7357. EWALD, J. A., DESOTELLE, J. A., WILDING, G. & JARRARD, D. F. 2010a. Therapy-Induced Senescence in Cancer. Journal of the National Cancer Institute. EWALD, J. A., DESOTELLE, J. A., WILDING, G. & JARRARD, D. F. 2010b. Therapy-Induced Senescence in Cancer. JNCI Journal of the National Cancer Institute, 102, 1536-1546. EWALD, J. A., PETERS, N., DESOTELLE, J. A., HOFFMANN, F. M. & JARRARD, D. F. 2009. A High- Throughput Method to Identify Novel Senescence-Inducing Compounds. Journal of Biomolecular Screening, 14, 853-858. EWALD, J. A. & JARRARD, D. F. 2012. Decreased Skp2 Expression Is Necessary but Not Sufficient for Therapy-Induced Senescence in Prostate Cancer. Translational Oncology, 5, 278-287. EWER, M. S. & EWER, S. M. 2015. Cardiotoxicity of anticancer treatments. Nat Rev Cardiol, advance online publication.

226

EXERTIER, P., JAVERZAT, S., WANG, B., FRANCO, M., HERBERT, J., PLATONOVA, N., WINANDY, M., PUJOL, N., NIVELLES, O., ORMENESE, S., GODARD, V., BECKER, J., BICKNELL, R., PINEAU, R., WILTING, J., BIKFALVI, A. & HAGEDORN, M. 2013. Impaired angiogenesis and tumor development by inhibition of the mitotic kinesin Eg5. Oncotarget, 4, 2302-2316. FAGAGNA, F. D. A. D., REAPER, P. M., CLAY-FARRACE, L., FIEGLER, H., CARR, P., VON ZGLINICKI, T., SARETZKI, G., CARTER, N. P. & JACKSON, S. P. 2003. A DNA damage checkpoint response in telomere-initiated senescence. Nature, 426, 194-198. FAN, S., EL-DEIRY, W. S., BAE, I., FREEMAN, J., JONDLE, D., BHATIA, K., FORNACE, A. J., MAGRATH, I., KOHN, K. W. & O'CONNOR, P. M. 1994. p53 Gene Mutations Are Associated with Decreased Sensitivity of Human Lymphoma Cells to DNA Damaging Agents. Cancer Research, 54, 5824- 5830. FENG, B., WANG, R. & CHEN, L.-B. 2012. MiR-100 resensitizes docetaxel-resistant human lung adenocarcinoma cells (SPC-A1) to docetaxel by targeting Plk1. Cancer letters, 317, 184-191. FERLAY, J., SHIN, H.-R., BRAY, F., FORMAN, D., MATHERS, C. & PARKIN, D. M. 2010. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. International Journal of Cancer, 127, 2893-2917. FERLINI, C., RASPAGLIO, G., MOZZETTI, S., DISTEFANO, M., FILIPPETTI, F., MARTINELLI, E., FERRANDINA, G., GALLO, D., RANELLETTI, F. O. & SCAMBIA, G. 2003. Bcl-2 Down-Regulation Is a Novel Mechanism of Paclitaxel Resistance. Molecular Pharmacology, 64, 51-58. FINK, D., NEBEL, S., AEBI, S., ZHENG, H., CENNI, B., NEHMÉ, A., CHRISTEN, R. D. & HOWELL, S. B. 1996. The Role of DNA Mismatch Repair in Platinum Drug Resistance. Cancer Research, 56, 4881-4886. FONTIJN, R. D., GOUD, B., ECHARD, A., JOLLIVET, F., VAN MARLE, J., PANNEKOEK, H. & HORREVOETS, A. J. G. 2001. The Human Kinesin-Like Protein RB6K Is under Tight Cell Cycle Control and Is Essential for Cytokinesis. Molecular and Cellular Biology, 21, 2944-2955. FRAGKOS, M. & BEARD, P. 2011. Mitotic Catastrophe Occurs in the Absence of Apoptosis in p53-Null Cells with a Defective G1 Checkpoint. PLoS One, 6, e22946. FRANCIS, R. E., MYATT, S. S., KROL, J., HARTMAN, J., PECK, B., MCGOVERN, U. B., WANG, J., GUEST, S. K., FILIPOVIC, A., GOJIS, O., PALMIERI, C., PESTON, D., SHOUSHA, S., YU, Q., SICINSKI, P., COOMBES, R. C. & LAM, E. W. F. 2009. FoxM1 is a downstream target and marker of HER2 overexpression in breast cancer. International journal of oncology, 35, 57-68. FURNARI, B., RHIND, N. & RUSSELL, P. 1997. Cdc25 Mitotic Inducer Targeted by Chk1 DNA Damage Checkpoint Kinase. Science, 277, 1495-1497. GANAPATHI, R. & GANAPATHI, M. K. 2013. Mechanisms Regulating Resistance to Inhibitors of Topoisomerase II. Frontiers in Pharmacology, 4. GANGULY, A., YANG, H., PEDROZA, M., BHATTACHARYA, R. & CABRAL, F. 2011. Mitotic Centromere- associated Kinesin (MCAK) Mediates Paclitaxel Resistance. Journal of Biological Chemistry, 286, 36378-36384. GASNEREAU, I., BOISSAN, M., MARGALL-DUCOS, G., COUCHY, G., WENDUM, D., BOURGAIN- GUGLIELMETTI, F., DESDOUETS, C., LACOMBE, M.-L., ZUCMAN-ROSSI, J. & SOBCZAK-THÉPOT, J. 2012. KIF20A mRNA and Its Product MKlp2 Are Increased During Hepatocyte Proliferation and Hepatocarcinogenesis. The American Journal of Pathology, 180, 131-140. GATEI, M., YOUNG, D., CEROSALETTI, K. M., DESAI-MEHTA, A., SPRING, K., KOZLOV, S., LAVIN, M. F., GATTI, R. A., CONCANNON, P. & KHANNA, K. 2000. ATM-dependent phosphorylation of nibrin in response to radiation exposure. Nat Genet, 25, 115-119. GEMENETZIDIS, E., BOSE, A., RIAZ, A. M., CHAPLIN, T., YOUNG, B. D., ALI, M., SUGDEN, D., THURLOW, J. K., CHEONG, S.-C., TEO, S.-H., WAN, H., WASEEM, A., PARKINSON, E. K., FORTUNE, F. & TEH, M.-T. 2009. FOXM1 Upregulation Is an Early Event in Human Squamous Cell Carcinoma and it Is Enhanced by Nicotine during Malignant Transformation. PLoS One, 4, e4849.

227

GEWIRTZ, D. 1999. A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochemical Pharmacology, 57, 727-741. GIANNAKAKOU, P., SACKETT, D. L., KANG, Y.-K., ZHAN, Z., BUTERS, J. T. M., FOJO, T. & PORUCHYNSKY, M. S. 1997. Paclitaxel-resistant Human Ovarian Cancer Cells Have Mutant β- Tubulins That Exhibit Impaired Paclitaxel-driven Polymerization. Journal of Biological Chemistry, 272, 17118-17125. GIFFIN, W., TORRANCE, H., RODDA, D. J., PREFONTAINE, G. G., POPE, L. & HACHE, R. J. G. 1996. Sequence-specific DNA binding by Ku autoantigen and its effects on transcription. Nature, 380, 265-268. GOBLE, S. & BEAR, H. D. 2003. Emerging role of taxanes in adjuvant and neoadjuvant therapy for breast cancer: The potential and the questions. Surgical Clinics of North America, 83, 943- 971. GOLDSTEIN, L. S. B. & PHILP, A. V. 1999. THE ROAD LESS TRAVELED1: Emerging Principles of Kinesin Motor Utilization. Annual Review of Cell and Developmental Biology, 15, 141-183. GORNSTEIN, E. & SCHWARZ, T. L. 2014. The paradox of paclitaxel neurotoxicity: Mechanisms and unanswered questions. Neuropharmacology, 76, Part A, 175-183. GOTTESMAN, M. M. 2002. Mechanisms of Cancer Drug Resistance. Annual Review of Medicine, 53, 615-627. GROTH-PEDERSEN, L., AITS, S., CORCELLE-TERMEAU, E., PETERSEN, N. H. T., NYLANDSTED, J. & JÄÄTTELÄ, M. 2012. Identification of Cytoskeleton-Associated Proteins Essential for Lysosomal Stability and Survival of Human Cancer Cells. PLoS One, 7, e45381. GRUNEBERG, U., NEEF, R., HONDA, R., NIGG, E. A. & BARR, F. A. 2004. Relocation of Aurora B from centromeres to the central spindle at the metaphase to anaphase transition requires MKlp2. The Journal of Cell Biology, 166, 167-172. GUANO, F., POURQUIER, P., TINELLI, S., BINASCHI, M., BIGIONI, M., ANIMATI, F., MANZINI, S., ZUNINO, F., KOHLHAGEN, G., POMMIER, Y. & CAPRANICO, G. 1999. Topoisomerase Poisoning Activity of Novel Disaccharide Anthracyclines. Molecular Pharmacology, 56, 77-84. GUSAROVA, G. A., WANG, I. C., MAJOR, M. L., KALINICHENKO, V. V., ACKERSON, T., PETROVIC, V. & COSTA, R. H. 2007. A cell-penetrating ARF peptide inhibitor of FoxM1 in mouse hepatocellular carcinoma treatment. The Journal of Clinical Investigation, 117, 99-111. GYÖRFFY, B., LANCZKY, A., EKLUND, A., DENKERT, C., BUDCZIES, J., LI, Q. & SZALLASI, Z. 2010. An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients. Breast Cancer Research and Treatment, 123, 725-731. HALASI, M. & GARTEL, A. L. 2013a. FOX(M1) NEWS – it is cancer. Molecular Cancer Therapeutics, 12, 245-254. HALASI, M. & GARTEL, A. L. 2013b. Targeting FOXM1 in cancer. Biochemical Pharmacology, 85, 644- 652. HARLEY, C. B., FUTCHER, A. B. & GREIDER, C. W. 1990. Telomeres shorten during ageing of human fibroblasts. Nature, 345, 458-460. HAYFLICK, L. 1965. The limited in vitro lifetime of human diploid cell strains. Experimental Cell Research, 37, 614-636. HEGDE, N. S., SANDERS, D. A., RODRIGUEZ, R. & BALASUBRAMANIAN, S. 2011. The transcription factor FOXM1 is a cellular target of the natural product thiostrepton. Nat Chem, 3, 725-731. HELLEDAY, T., PETERMANN, E., LUNDIN, C., HODGSON, B. & SHARMA, R. A. 2008. DNA repair pathways as targets for cancer therapy. Nat Rev Cancer, 8, 193-204. HILL, E., CLARKE, M. & BARR, F. A. 2000. The Rab6-binding kinesin, Rab6-KIFL, is required for cytokinesis. EMBO Journal, 19, 5711-5719. HIROKAWA, N., NODA, Y. & OKADA, Y. 1998. Kinesin and dynein superfamily proteins in organelle transport and cell division. Current Opinion in Cell Biology, 10, 60-73.

228

HO, J. R., CHAPEAUBLANC, E., KIRKWOOD, L., NICOLLE, R., BENHAMOU, S., LEBRET, T., ALLORY, Y., SOUTHGATE, J., RADVANYI, F. & GOUD, B. 2012. Deregulation of Rab and Rab Effector Genes in Bladder Cancer. PLoS One, 7, e39469. HOLOHAN, C., VAN SCHAEYBROECK, S., LONGLEY, D. B. & JOHNSTON, P. G. 2013. Cancer drug resistance: an evolving paradigm. Nat Rev Cancer, 13, 714-726. HOREJSI, Z., FALCK, J., BAKKENIST, C. J., KASTAN, M. B., LUKAS, J. & BARTEK, J. 2004. Distinct functional domains of Nbs1 modulate the timing and magnitude of ATM activation after low doses of ionizing radiation. Oncogene, 23, 3122-3127. HOSOYA, N. & MIYAGAWA, K. 2014. Targeting DNA damage response in cancer therapy. Cancer Science, 105, 370-388. HOU, X., LI, Z., HUANG, W., LI, J., STAIGER, C., KUANG, S., RATLIFF, T. & LIU, X. 2013. Plk1-dependent microtubule dynamics promotes androgen receptor signaling in prostate cancer. The Prostate, 73, 1352-1363. HOWELL, A., ANDERSON, A., CLARKE, R., DUFFY, S., EVANS, D., GARCIA-CLOSAS, M., GESCHER, A., KEY, T., SAXTON, J. & HARVIE, M. 2014. Risk determination and prevention of breast cancer. Breast Cancer Research, 16, 446. HSU, D. S.-S., CHANG, S.-Y., LIU, C.-J., TZENG, C.-H., WU, K.-J., KAO, J.-Y. & YANG, M.-H. 2010. Identification of increased NBS1 expression as a prognostic marker of squamous cell carcinoma of the oral cavity. Cancer Science, 101, 1029-1037. HUCK, J. J., ZHANG, M., MCDONALD, A., BOWMAN, D., HOAR, K. M., STRINGER, B., ECSEDY, J., MANFREDI, M. G. & HYER, M. L. 2010. MLN8054, an Inhibitor of Aurora A Kinase, Induces Senescence in Human Tumor Cells Both In vitro and In vivo. Molecular Cancer Research, 8, 373-384. HUI, M. C., CHAN, K., LUK, J., LEE, N., CHUNG, Y., CHEUNG, L. M., SRIVASTAVA, G., TSAO, S., TANG, J. & LAW, S. 2012. Cytoplasmic Forkhead Box M1 (FoxM1) in Esophageal Squamous Cell Carcinoma Significantly Correlates with Pathological Disease Stage. World Journal of Surgery, 36, 90-97. ITAHANA, K., CAMPISI, J. & DIMRI, G. P. 2007. Methods to Detect Biomarkers of Cellular Senescence. ITAHANA, K., ITAHANA, Y. & DIMRI, G. P. 2013. Colorimetric Detection of Senescence-Associated β Galactosidase. Methods in molecular biology (Clifton, N.J.), 965, 143-156. IVASHKEVICH, A., REDON, C. E., NAKAMURA, A. J., MARTIN, R. F. & MARTIN, O. A. 2012. USE OF THE γ-H2AX ASSAY TO MONITOR DNA DAMAGE AND REPAIR IN TRANSLATIONAL CANCER RESEARCH. Cancer letters, 327, 123-133. IWAKIRI, T., OKUMURA, M., HIDAKA, M., KUMAGAI, Y., ICHIHARA, E., KAWANO, Y. & ARIMORI, K. 2008. Inhibition of carrier-mediated uptake of epirubicin reduces cytotoxicity in primary culture of rat hepatocytes. Journal of Applied Toxicology, 28, 329-336. JACKSON, S. P. & BARTEK, J. 2009. The DNA-damage response in human biology and disease. Nature, 461, 1071-1078. JAIYESIMI, I. A., BUZDAR, A. U., DECKER, D. A. & HORTOBAGYI, G. N. 1995. Use of tamoxifen for breast cancer: twenty-eight years later. Journal of Clinical Oncology, 13, 513-29. JALAL, S., EARLEY, J. N. & TURCHI, J. J. 2011. DNA Repair: From Genome Maintenance to Biomarker and Therapeutic Target. Clinical Cancer Research, 17, 6973-6984. JIANG, L., SIU, M. K. Y., WONG, O. G. W., TAM, K.-F., LU, X., LAM, E. W.-F., NGAN, H. Y. S., LE, X.-F., WONG, E. S. Y., MONTEIRO, L. J., CHAN, H.-Y. & CHEUNG, A. N. Y. 2011. iASPP and Chemoresistance in Ovarian Cancers: Effects on Paclitaxel-Mediated Mitotic Catastrophe. Clinical Cancer Research, 17, 6924-6933. JIRAWATNOTAI, S., HU, Y., MICHOWSKI, W., ELIAS, J. E., BECKS, L., BIENVENU, F., ZAGOZDZON, A., GOSWAMI, T., WANG, Y. E., CLARK, A. B., KUNKEL, T. A., VAN HARN, T., XIA, B., CORRELL, M., QUACKENBUSH, J., LIVINGSTON, D. M., GYGI, S. P. & SICINSKI, P. 2011. A function for cyclin D1 in DNA repair uncovered by protein interactome analyses in human cancers. Nature, 474, 230-234.

229

JOHNSON, R. D. & JASIN, M. 2000. Sister chromatid gene conversion is a prominent double-strand break repair pathway in mammalian cells. The EMBO Journal, 19, 3398-3407. JORDAN, M. A., WENDELL, K., GARDINER, S., BRENT DERRY, W., COPP, H. & WILSON, L. 1996. Mitotic Block Induced in HeLa Cells by Low Concentrations of Paclitaxel (Taxol) Results in Abnormal Mitotic Exit and Apoptotic Cell Death. Cancer Research, 56, 816-825. JUCHA, A., WEGIEREK-CIUK, A., KOZA, Z., LISOWSKA, H., WOJCIK, A., WOJEWODZKA, M. & LANKOFF, A. 2010. FociCounter: A freely available PC programme for quantitative and qualitative analysis of gamma-H2AX foci. Mutation Research/Genetic Toxicology and Environmental Mutagenesis, 696, 16-20. KAHLEM, P., XF, RKEN, B. & SCHMITT, C. A. 2004. Cellular senescence in cancer treatment: friend or foe? The Journal of Clinical Investigation, 113, 169-174. KALINICHENKO, V. V., MAJOR, M. L., WANG, X., PETROVIC, V., KUECHLE, J., YODER, H. M., DENNEWITZ, M. B., SHIN, B., DATTA, A., RAYCHAUDHURI, P. & COSTA, R. H. 2004. Foxm1b transcription factor is essential for development of hepatocellular carcinomas and is negatively regulated by the p19ARF tumor suppressor. Genes & Development, 18, 830-850. KANG, J., FERGUSON, D., SONG, H., BASSING, C., ECKERSDORFF, M., ALT, F. W. & XU, Y. 2005. Functional Interaction of H2AX, NBS1, and p53 in ATM-Dependent DNA Damage Responses and Tumor Suppression. Molecular and Cellular Biology, 25, 661-670. KAPSE-MISTRY, S., GOVENDER, T., SRIVASTAVA, R. & YERGERI, M. 2014. Nanodrug Delivery in Reversing Multidrug Resistance in Cancer Cells. Frontiers in Pharmacology, 5. KARADEDOU, C. T., GOMES, A. R., CHEN, J., PETKOVIC, M., HO, K. K., ZWOLINSKA, A. K., FELTES, A., WONG, S. Y., CHAN, K. Y. K., CHEUNG, Y. N., TSANG, J. W. H., BROSENS, J. J., KHOO, U. S. & LAM, E. W. F. 2012. FOXO3a represses VEGF expression through FOXM1-dependent and - independent mechanisms in breast cancer. Oncogene, 31, 1845-1858. KARIMI-BUSHERI, F., RASOULI-NIA, A., MACKEY, J. R. & WEINFELD, M. 2010. Senescence evasion by MCF-7 human breast tumor-initiating cells. Breast Cancer Research : BCR, 12, R31-R31. KELSEY, J. L. & BERKOWITZ, G. S. 1988. Breast Cancer Epidemiology. Cancer Research, 48, 5615-5623. KHANNA, K. K. & JACKSON, S. P. 2001. DNA double-strand breaks: signaling, repair and the cancer connection. Nat Genet, 27, 247-254. KHONGKOW, M., OLMOS, Y., GONG, C., GOMES, A. R., MONTEIRO, L. J., YAGÜE, E., CAVACO, T. B., KHONGKOW, P., MAN, E. P. S., LAOHASINNARONG, S., KOO, C.-Y., HARADA-SHOJI, N., TSANG, J. W.-H., COOMBES, R. C., SCHWER, B., KHOO, U.-S. & LAM, E. W.-F. 2013. SIRT6 modulates paclitaxel and epirubicin resistance and survival in breast cancer. Carcinogenesis, 34, 1476-1486. KHONGKOW, P., GOMES, A. R., GONG, C., MAN, E. P. S., TSANG, J. W. H., ZHAO, F., MONTEIRO, L. J., COOMBES, R. C., MEDEMA, R. H., KHOO, U. S. & LAM, E. W. F. 2015. Paclitaxel targets FOXM1 to regulate KIF20A in mitotic catastrophe and breast cancer paclitaxel resistance. Oncogene. KHONGKOW, P., KARUNARATHNA, U., KHONGKOW, M., GONG, C., GOMES, A. R., YAGUE, E., MONTEIRO, L. J., KONGSEMA, M., ZONA, S., MAN, E. P. S., TSANG, J. W. H., COOMBES, R. C., WU, K. J., KHOO, U. S., MEDEMA, R. H., FREIRE, R. & LAM, E. W. F. 2014. FOXM1 targets NBS1 to regulate DNA damage-induced senescence and epirubicin resistance. Oncogene, 33, 4144-4155. KHONGKOW, P., KARUNARATHNA, U. & LAM, E. 2013. Abstract A92: The role of FOXM1 and NBS1 in DNA damage-induced senescence and epirubicin resistance. Molecular Cancer Therapeutics, 12, A92. KHONGKOW, P., KARUNARATHNA, H. U. & LAM, E. W. 2012 The Role of FOXM1 and NBS1 in DNA Double Strand Breaks Repair and Epirubicin Resistance. European Journal of Cancer, 48, S173, 728. KOBAYASHI, J., ANTOCCIA, A., TAUCHI, H., MATSUURA, S. & KOMATSU, K. 2004. NBS1 and its functional role in the DNA damage response. DNA Repair, 3, 855-861.

230

KOO, C.-Y., MUIR, K. W. & LAM, E. W. F. 2012. FOXM1: From cancer initiation to progression and treatment. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms, 1819, 28-37. KOPS, G. J. P. L., WEAVER, B. A. A. & CLEVELAND, D. W. 2005. On the road to cancer: aneuploidy and the mitotic checkpoint. Nat Rev Cancer, 5, 773-785. KORVER, W., ROOSE, J. & CLEVERS, H. 1997a. The winged-helix transcription factor Trident is expressed in cycling cells. Nucleic Acids Research, 25, 1715-1719. KORVER, W., ROOSE, J., HEINEN, K., WEGHUIS, D. O., DE BRUIJN, D., VAN KESSEL, A. G. & CLEVERS, H. 1997b. The HumanTRIDENT/HFH-11/FKHL16Gene: Structure, Localization, and Promoter Characterization. Genomics, 46, 435-442. KREJCI, L., ALTMANNOVA, V., SPIREK, M. & ZHAO, X. 2012. Homologous recombination and its regulation. Nucleic Acids Research, 40, 5795-5818. KRÖGER, N., ACHTERRATH, W., HEGEWISCH-BECKER, S., MROSS, K. & ZANDER, A. R. 1999. Current options in treatment of anthracycline-resistant breast cancer. Cancer Treatment Reviews, 25, 279-291. KUILMAN, T., MICHALOGLOU, C., MOOI, W. J. & PEEPER, D. S. 2010. The essence of senescence. Genes & Development, 24, 2463-2479. KURZ, E. U. & LEES-MILLER, S. P. 2004. DNA damage-induced activation of ATM and ATM-dependent signaling pathways. DNA Repair, 3, 889-900. KWOK, J. M.-M., MYATT, S. S., MARSON, C. M., COOMBES, R. C., CONSTANTINIDOU, D. & LAM, E. W.- F. 2008. Thiostrepton selectively targets breast cancer cells through inhibition of forkhead box M1 expression. Molecular Cancer Therapeutics, 7, 2022-2032. KWOK, J. M.-M., PECK, B., MONTEIRO, L. J., SCHWENEN, H. D. C., MILLOUR, J., COOMBES, R. C., MYATT, S. S. & LAM, E. W.-F. 2010a. FOXM1 Confers Acquired Cisplatin Resistance in Breast Cancer Cells. Molecular Cancer Research. KWOK, J. M. M., PECK, B., MONTEIRO, L. J., SCHWENEN, H. D. C., MILLOUR, J., COOMBES, R. C., MYATT, S. S. & LAM, E. W. F. 2010b. FOXM1 confers acquired cisplatin resistance in breast cancer cells. Molecular cancer research : MCR, 8, 24-34. LAI, F., FERNALD, A. A., ZHAO, N. & LE BEAU, M. M. 2000. cDNA cloning, expression pattern, genomic structure and chromosomal location of RAB6KIFL, a human kinesin-like gene. Gene, 248, 117-125. LAM, A. K. Y., NGAN, A. W. L., LEUNG, M.-H., KWOK, D. C. T., LIU, V. W. S., CHAN, D. W., LEUNG, W. Y. & YAO, K.-M. 2013. FOXM1b, which is present at elevated levels in cancer cells, has a greater transforming potential than FOXM1c. Frontiers in Oncology, 3, 11. LAM, E. W. F., BROSENS, J. J., GOMES, A. R. & KOO, C.-Y. 2013. Forkhead box proteins: tuning forks for transcriptional harmony. Nat Rev Cancer, 13, 482-495. LANDT, S. G., MARINOV, G. K., KUNDAJE, A., KHERADPOUR, P., PAULI, F., BATZOGLOU, S., BERNSTEIN, B. E., BICKEL, P., BROWN, J. B., CAYTING, P., CHEN, Y., DESALVO, G., EPSTEIN, C., FISHER-AYLOR, K. I., EUSKIRCHEN, G., GERSTEIN, M., GERTZ, J., HARTEMINK, A. J., HOFFMAN, M. M., IYER, V. R., JUNG, Y. L., KARMAKAR, S., KELLIS, M., KHARCHENKO, P. V., LI, Q., LIU, T., LIU, X. S., MA, L., MILOSAVLJEVIC, A., MYERS, R. M., PARK, P. J., PAZIN, M. J., PERRY, M. D., RAHA, D., REDDY, T. E., ROZOWSKY, J., SHORESH, N., SIDOW, A., SLATTERY, M., STAMATOYANNOPOULOS, J. A., TOLSTORUKOV, M. Y., WHITE, K. P., XI, S., FARNHAM, P. J., LIEB, J. D., WOLD, B. J. & SNYDER, M. 2012. ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Research, 22, 1813-1831. LAOUKILI, J., ALVAREZ-FERNANDEZ, M., STAHL, M. & MEDEMA, R. H. 2008a. FoxM1 is degraded at mitotic exit in a Cdh1-dependent manner. Cell Cycle, 7, 2720-2726. LAOUKILI, J., ALVAREZ, M., MEIJER, L. A. T., STAHL, M., MOHAMMED, S., KLEIJ, L., HECK, A. J. R. & MEDEMA, R. H. 2008b. Activation of FoxM1 during G2 Requires Cyclin A/Cdk-Dependent Relief of Autorepression by the FoxM1 N-Terminal Domain. Molecular and Cellular Biology, 28, 3076-3087.

231

LAOUKILI, J., KOOISTRA, M. R. H., BRAS, A., KAUW, J., KERKHOVEN, R. M., MORRISON, A., CLEVERS, H. & MEDEMA, R. H. 2005. FoxM1 is required for execution of the mitotic programme and chromosome stability. Nat Cell Biol, 7, 126-136. LAOUKILI, J., STAHL, M. & MEDEMA, R. H. 2007. FoxM1: At the crossroads of ageing and cancer. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer, 1775, 92-102. LARA-GONZALEZ, P., WESTHORPE, FREDERICK G. & TAYLOR, STEPHEN S. 2012. The Spindle Assembly Checkpoint. Current Biology, 22, R966-R980. LEE, J.-H. & PAULL, T. T. 2005. ATM Activation by DNA Double-Strand Breaks Through the Mre11- Rad50-Nbs1 Complex. Science, 308, 551-554. LEE, J. H. & PAULL, T. T. 2007. Activation and regulation of ATM kinase activity in response to DNA double-strand breaks. Oncogene, 26, 7741-7748. LEUNG, T. W. C., LIN, S. S. W., TSANG, A. C. C., TONG, C. S. W., CHING, J. C. Y., LEUNG, W. Y., GIMLICH, R., WONG, G. G. & YAO, K.-M. Over-expression of FoxM1 stimulates cyclin B1 expression. FEBS Letters, 507, 59-66. LI, J., WANG, Y., LUO, J., FU, Z., YING, J., YU, Y. & YU, W. 2012. MiR-134 inhibits epithelial to mesenchymal transition by targeting FOXM1 in non-small cell lung cancer cells. FEBS Letters, 586, 3761-3765. LI, Q., ZHANG, N., JIA, Z., LE, X., DAI, B., WEI, D., HUANG, S., TAN, D. & XIE, K. 2009. Critical Role and Regulation of Transcription Factor FoxM1 in Human Gastric Cancer Angiogenesis and Progression. Cancer Research, 69, 3501-3509. LI, X., YAO, R., YUE, L., QIU, W., QI, W., LIU, S., YAO, Y. & LIANG, J. 2014. FOXM1 mediates resistance to docetaxel in gastric cancer via up-regulating Stathmin. Journal of Cellular and Molecular Medicine, 18, 811-823. LIM, D.-S., KIM, S.-T., XU, B., MASER, R. S., LIN, J., PETRINI, J. H. J. & KASTAN, M. B. 2000. ATM phosphorylates p95/nbs1 in an S-phase checkpoint pathway. Nature, 404, 613-617. LITTLER, D. R., ALVAREZ-FERNÁNDEZ, M., STEIN, A., HIBBERT, R. G., HEIDEBRECHT, T., ALOY, P., MEDEMA, R. H. & PERRAKIS, A. 2010. Structure of the FoxM1 DNA-recognition domain bound to a promoter sequence. Nucleic Acids Research, 38, 4527-4538. LIU, M., DAI, B., KANG, S.-H., BAN, K., HUANG, F.-J., LANG, F. F., ALDAPE, K. D., XIE, T.-X., PELLOSKI, C. E., XIE, K., SAWAYA, R. & HUANG, S. 2006. FoxM1B Is Overexpressed in Human Glioblastomas and Critically Regulates the Tumorigenicity of Glioma Cells. Cancer Research, 66, 3593-3602. LONGLEY, D. B. & JOHNSTON, P. G. 2005. Molecular mechanisms of drug resistance. The Journal of Pathology, 205, 275-292. LORD, C. J. & ASHWORTH, A. 2012. The DNA damage response and cancer therapy. Nature, 481, 287- 294. LU, Y., ZI, X., ZHAO, Y., MASCARENHAS, D. & POLLAK, M. 2001. Insulin-Like Growth Factor-I Receptor Signaling and Resistance to Trastuzumab (Herceptin). Journal of the National Cancer Institute, 93, 1852-1857. LUNDGREN, K., WALWORTH, N., BOOHER, R., DEMBSKI, M., KIRSCHNER, M. & BEACH, D. 1991. mik1 and wee1 cooperate in the inhibitory tyrosine phosphorylation of cdc2. Cell, 64, 1111-1122. MATSUOKA, S., HUANG, M. & ELLEDGE, S. J. 1998. Linkage of ATM to Cell Cycle Regulation by the Chk2 Protein Kinase. Science, 282, 1893-1897. MYATT, S. S. & LAM, E. W. F. 2007. The emerging roles of forkhead box (Fox) proteins in cancer. Nat Rev Cancer, 7, 847-859. MCGOVERN, U. B., FRANCIS, R. E., PECK, B., GUEST, S. K., WANG, J., MYATT, S. S., KROL, J., KWOK, J. M.-M., POLYCHRONIS, A., COOMBES, R. C. & LAM, E. W.-F. 2009. Gefitinib (Iressa) represses FOXM1 expression via FOXO3a in breast cancer. Molecular Cancer Therapeutics, 8, 582-591. MICHAELIS, M., ROTHWEILER, F., AGHA, B., BARTH, S., VOGES, Y., LÖSCHMANN, N., VON DEIMLING, A., BREITLING, R., WILHELM DOERR, H., RÖDEL, F., SPEIDEL, D. & CINATL, J. 2012. Human

232

neuroblastoma cells with acquired resistance to the p53 activator RITA retain functional p53 and sensitivity to other p53 activating agents. Cell Death & Disease, 3, e294. MICHALOGLOU, C., VREDEVELD, L. C. W., SOENGAS, M. S., DENOYELLE, C., KUILMAN, T., VAN DER HORST, C. M. A. M., MAJOOR, D. M., SHAY, J. W., MOOI, W. J. & PEEPER, D. S. 2005. BRAFE600-associated senescence-like cell cycle arrest of human naevi. Nature, 436, 720-724. MIKI, Y., SWENSEN, J., SHATTUCK-EIDENS, D., FUTREAL, P., HARSHMAN, K., TAVTIGIAN, S., LIU, Q., COCHRAN, C., BENNETT, L., DING, W. & ET, A. 1994. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science, 266, 66-71. MILLION WOMEN STUDY, C. & THE COLLABORATORS ARE LISTED ON, P. 2002. Patterns of use of hormone replacement therapy in one million women in Britain, 1996–2000. BJOG: An International Journal of Obstetrics & Gynaecology, 109, 1319-1330. MILLOUR, J., DE OLANO, N., HORIMOTO, Y., MONTEIRO, L. J., LANGER, J. K., ALIGUE, R., HAJJI, N. & LAM, E. W.-F. 2011. ATM and p53 Regulate FOXM1 Expression via E2F in Breast Cancer Epirubicin Treatment and Resistance. Molecular Cancer Therapeutics, 10, 1046-1058. MINOTTI, A. M., BARLOW, S. B. & CABRAL, F. 1991. Resistance to antimitotic drugs in Chinese hamster ovary cells correlates with changes in the level of polymerized tubulin. Journal of Biological Chemistry, 266, 3987-3994. MINOTTI, G., MENNA, P., SALVATORELLI, E., CAIRO, G. & GIANNI, L. 2004. Anthracyclines: Molecular Advances and Pharmacologic Developments in Antitumor Activity and Cardiotoxicity. Pharmacological Reviews, 56, 185-229. MONTEIRO, L. J., KHONGKOW, P., KONGSEMA, M., MORRIS, J. R., MAN, C., WEEKES, D., KOO, C. Y., GOMES, A. R., PINTO, P. H., VARGHESE, V., KENNY, L. M., CHARLES COOMBES, R., FREIRE, R., MEDEMA, R. H. & LAM, E. W. F. 2013. The Forkhead Box M1 protein regulates BRIP1 expression and DNA damage repair in epirubicin treatment. Oncogene, 32, 4634-4645. MONTERO, A., FOSSELLA, F., HORTOBAGYI, G. & VALERO, V. 2005. Docetaxel for treatment of solid tumours: a systematic review of clinical data. The Lancet Oncology, 6, 229-239. MORENO-ASPITIA, A. & PEREZ, E. A. 2009. Anthracycline- and/or taxane-resistant breast cancer: Results of a literature review to determine the clinical challenges and current treatment trends. Clinical Therapeutics, 31, 1619-1640. MORRIS, J. R., BOUTELL, C., KEPPLER, M., DENSHAM, R., WEEKES, D., ALAMSHAH, A., BUTLER, L., GALANTY, Y., PANGON, L., KIUCHI, T., NG, T. & SOLOMON, E. 2009. The SUMO modification pathway is involved in the BRCA1 response to genotoxic stress. Nature, 462, 886-890. MORSE, D. L., GRAY, H., PAYNE, C. M. & GILLIES, R. J. 2005. Docetaxel induces cell death through mitotic catastrophe in human breast cancer cells. Molecular Cancer Therapeutics, 4, 1495- 1504. MOYNAHAN, M. E. & JASIN, M. 2010. Mitotic homologous recombination maintains genomic stability and suppresses tumorigenesis. Nat Rev Mol Cell Biol, 11, 196-207. MURRAY, S., BRIASOULIS, E., LINARDOU, H., BAFALOUKOS, D. & PAPADIMITRIOU, C. 2012. Taxane resistance in breast cancer: Mechanisms, predictive biomarkers and circumvention strategies. Cancer Treatment Reviews, 38, 890-903. MUSACCHIO, A. & SALMON, E. D. 2007. The spindle-assembly checkpoint in space and time. Nat Rev Mol Cell Biol, 8, 379-393. MYATT, S. S., KONGSEMA, M., MAN, C. W. Y., KELLY, D. J., GOMES, A. R., KHONGKOW, P., KARUNARATHNA, U., ZONA, S., LANGER, J. K., DUNSBY, C. W., COOMBES, R. C., FRENCH, P. M., BROSENS, J. J. & LAM, E. W. F. 2014. SUMOylation inhibits FOXM1 activity and delays mitotic transition. Oncogene, 33, 4316-4329. MYATT, S. S. & LAM, E. W. F. 2007. The emerging roles of forkhead box (Fox) proteins in cancer. Nat Rev Cancer, 7, 847-859. NAKAMURA, S., HIRANO, I., OKINAKA, K., TAKEMURA, T., YOKOTA, D., ONO, T., SHIGENO, K., SHIBATA, K., FUJISAWA, S. & OHNISHI, K. 2010. The FOXM1 transcriptional factor promotes

233

the proliferation of leukemia cells through modulation of cell cycle progression in acute myeloid leukemia. Carcinogenesis, 31, 2012-2021. NARDELLA, C., CLOHESSY, J. G., ALIMONTI, A. & PANDOLFI, P. P. 2011. Pro-senescence therapy for cancer treatment. Nat Rev Cancer, 11, 503-511. NEEF, R., KLEIN, U. R., KOPAJTICH, R. & BARR, F. A. 2006. Cooperation between Mitotic Kinesins Controls the Late Stages of Cytokinesis. Current Biology, 16, 301-307. NEEF, R., PREISINGER, C., SUTCLIFFE, J., KOPAJTICH, R., NIGG, E. A., MAYER, T. U. & BARR, F. A. 2003. Phosphorylation of mitotic kinesin-like protein 2 by polo-like kinase 1 is required for cytokinesis. The Journal of Cell Biology, 162, 863-876. OHNISHI, K., SCURIC, Z., SCHIESTL, R. H., OKAMOTO, N., TAKAHASHI, A. & OHNISHI, T. 2006. siRNA Targeting NBS1 or XIAP Increases Radiation Sensitivity of Human Cancer Cells Independent of TP53 Status. Radiation Research, 166, 454-462. OKADA, H. & MAK, T. W. 2004. Pathways of apoptotic and non-apoptotic death in tumour cells. Nat Rev Cancer, 4, 592-603. ORR, G. A., VERDIER-PINARD, P., MCDAID, H. & HORWITZ, S. B. 2003. Mechanisms of Taxol resistance related to microtubules. Oncogene, 22, 7280-7295. OVERALL, C. M. & KLEIFELD, O. 2006. Validating matrix metalloproteinases as drug targets and anti- targets for cancer therapy. Nat Rev Cancer, 6, 227-239. PANDIT, B. & GARTEL, A. L. 2010. New potential anti-cancer agents synergize with bortezomib and ABT-737 against prostate cancer. Prostate, 70, 825-833. PARK, H. J., GUSAROVA, G., WANG, Z., CARR, J. R., LI, J., KIM, K. H., QIU, J., PARK, Y. D., WILLIAMSON, P. R., HAY, N., TYNER, A. L., LAU, L. F., COSTA, R. H. & RAYCHAUDHURI, P. 2010. Deregulation of FoxM1b leads to tumour metastasis. EMBO Molecular Medicine, 3, 21-34. PARK, H. J., GUSAROVA, G., WANG, Z., CARR, J. R., LI, J., KIM, K. H., QIU, J., PARK, Y. D., WILLIAMSON, P. R., HAY, N., TYNER, A. L., LAU, L. F., COSTA, R. H. & RAYCHAUDHURI, P. 2011. Deregulation of FoxM1b leads to tumour metastasis. PARK, H. J., WANG, Z., COSTA, R. H., TYNER, A., LAU, L. F. & RAYCHAUDHURI, P. 2007. An N-terminal inhibitory domain modulates activity of FoxM1 during cell cycle. Oncogene, 27, 1696-1704. PARK, Y.-Y., JUNG, S. Y., JENNINGS, N. B., RODRIGUEZ-AGUAYO, C., PENG, G., LEE, S.-R., KIM, S. B., KIM, K., LEEM, S.-H., LIN, S.-Y., LOPEZ-BERESTEIN, G., SOOD, A. K. & LEE, J.-S. 2012. FOXM1 mediates Dox resistance in breast cancer by enhancing DNA repair. Carcinogenesis, 33, 1843-1853. PAZOLLI, E., ALSPACH, E., MILCZAREK, A., PRIOR, J., PIWNICA-WORMS, D. & STEWART, S. A. 2012. Chromatin remodeling underlies the senescence-associated secretory phenotype of tumor stromal fibroblasts that supports cancer progression. Cancer Research, 72, 2251-2261. PEREZ, R. P. 1998. Cellular and molecular determinants of cisplatin resistance. European Journal of Cancer, 34, 1535-1542. PIERCE, A. J., JOHNSON, R. D., THOMPSON, L. H. & JASIN, M. 1999. XRCC3 promotes homology- directed repair of DNA damage in mammalian cells. Genes & Development, 13, 2633-2638. PIETENPOL, J. A. & STEWART, Z. A. 2002. Cell cycle checkpoint signaling:: Cell cycle arrest versus apoptosis. Toxicology, 181–182, 475-481. PILARSKY, C., WENZIG, M., SPECHT, T., SAEGER, H. D. & GRÜTZMANN, R. 2004. Identification and Validation of Commonly Overexpressed Genes in Solid Tumors by Comparison of Microarray Data. Neoplasia, 6, 744-750. PISANI, P., BRAY, F. & PARKIN, D. M. 2002. Estimates of the world-wide prevalence of cancer for 25 sites in the adult population. International Journal of Cancer, 97, 72-81. PRILLER, M., PÖSCHL, J., ABRÃO, L., VON BUEREN, A. O., CHO, Y.-J., RUTKOWSKI, S., KRETZSCHMAR, H. A. & SCHÜLLER, U. 2011. Expression of FoxM1 Is Required for the Proliferation of Medulloblastoma Cells and Indicates Worse Survival of Patients. Clinical Cancer Research, 17, 6791-6801.

234

RADHAKRISHNAN, S. K., BHAT, U. G., HUGHES, D. E., WANG, I.-C., COSTA, R. H. & GARTEL, A. L. 2006. Identification of a Chemical Inhibitor of the Oncogenic Transcription Factor Forkhead Box M1. Cancer Research, 66, 9731-9735. RATH, O. & KOZIELSKI, F. 2012. Kinesins and cancer. Nat Rev Cancer, 12, 527-539. RAYCHAUDHURI, P. & PARK, H. J. 2011. FoxM1: A Master Regulator of Tumor Metastasis. Cancer Research, 71, 4329-4333. REBBAA, A., ZHENG, X., CHOU, P. M. & MIRKIN, B. L. 2003. Caspase inhibition switches doxorubicin- induced apoptosis to senescence. Oncogene, 22, 2805-2811. RHIND, N. & RUSSELL, P. 2012. Signaling Pathways that Regulate Cell Division. Cold Spring Harbor Perspectives in Biology, 4. RICCI, M. S. & ZONG, W.-X. 2006. Chemotherapeutic Approaches for Targeting Cell Death Pathways. The oncologist, 11, 342-357. RINK, L., SLUPIANEK, A., STOKLOSA, T., NIEBOROWSKA-SKORSKA, M., URBANSKA, K., SEFERYNSKA, I., REISS, K. & SKORSKI, T. 2007. Enhanced phosphorylation of Nbs1, a member of DNA repair/checkpoint complex Mre11-RAD50-Nbs1, can be targeted to increase the efficacy of imatinib mesylate against BCR/ABL-positive leukemia cells. Blood, 110, 651-660. ROBERSON, R. S., KUSSICK, S. J., VALLIERES, E., CHEN, S.-Y. J. & WU, D. Y. 2005. Escape from Therapy- Induced Accelerated Cellular Senescence in p53-Null Lung Cancer Cells and in Human Lung Cancers. Cancer Research, 65, 2795-2803. RODIER, F., COPPE, J.-P., PATIL, C. K., HOEIJMAKERS, W. A. M., MUNOZ, D. P., RAZA, S. R., FREUND, A., CAMPEAU, E., DAVALOS, A. R. & CAMPISI, J. 2009a. Persistent DNA damage signalling triggers senescence-associated inflammatory cytokine secretion. Nat Cell Biol, 11, 973-979. RODIER, F., COPPÉ, J.-P., PATIL, C. K., HOEIJMAKERS, W. A. M., MUÑOZ, D. P., RAZA, S. R., FREUND, A., CAMPEAU, E., DAVALOS, A. R. & CAMPISI, J. 2009b. Persistent DNA damage signaling triggers senescence-associated inflammatory cytokine secretion. Nature cell biology, 11, 973-979. ROGAKOU, E. P., PILCH, D. R., ORR, A. H., IVANOVA, V. S. & BONNER, W. M. 1998. DNA Double- stranded Breaks Induce Histone H2AX Phosphorylation on Serine 139. Journal of Biological Chemistry, 273, 5858-5868. RONINSON, I. B., BROUDE, E. V. & CHANG, B.-D. 2001. If not apoptosis, then what? Treatment- induced senescence and mitotic catastrophe in tumor cells. Drug Resistance Updates, 4, 303- 313. ROUZIER, R., RAJAN, R., WAGNER, P., HESS, K. R., GOLD, D. L., STEC, J., AYERS, M., ROSS, J. S., ZHANG, P., BUCHHOLZ, T. A., KUERER, H., GREEN, M., ARUN, B., HORTOBAGYI, G. N., SYMMANS, W. F. & PUSZTAI, L. 2005. Microtubule-associated protein tau: A marker of paclitaxel sensitivity in breast cancer. Proceedings of the National Academy of Sciences of the United States of America, 102, 8315-8320. ROVILLAIN, E., MANSFIELD, L., CAETANO, C., ALVAREZ-FERNANDEZ, M., CABALLERO, O. L., MEDEMA, R. H., HUMMERICH, H. & JAT, P. S. 2011. Activation of nuclear factor-kappa B signalling promotes cellular senescence. Oncogene, 30, 2356-2366. RUSSELL, P. & NURSE, P. 1987. Negative regulation of mitosis by wee1+, a gene encoding a protein kinase homolog. Cell, 49, 559-567. S.C. WILSON, M., J. BROSENS, J., D.C. SCHWENEN, H. & W.-F. LAM, E. 2011. FOXO and FOXM1 in Cancer: The FOXO-FOXM1 Axis Shapes the Outcome of Cancer Chemotherapy. Current Drug Targets, 12, 1256-1266. SAAR, K., CHRZANOWSKA, K. H., STUMM, M., JUNG, M., NÜRNBERG, G., WIENKER, T. F., SEEMANOVÁ, E., WEGNER, R. D., REIS, A. & SPERLING, K. 1997. The gene for the ataxia- telangiectasia variant, Nijmegen breakage syndrome, maps to a 1-cM interval on chromosome 8q21. American Journal of Human Genetics, 60, 605-610. SAITO, Y., FUJIMOTO, H. & KOBAYASHI, J. 2013. Role of NBS1 in DNA damage response and its relationship with cancer development. Translational Cancer Research, 2, 178-189.

235

SANCHEZ, Y., WONG, C., THOMA, R. S., RICHMAN, R., WU, Z., PIWNICA-WORMS, H. & ELLEDGE, S. J. 1997. Conservation of the Chk1 Checkpoint Pathway in Mammals: Linkage of DNA Damage to Cdk Regulation Through Cdc25. Science, 277, 1497-1501. SARKARIA, J. N., TIBBETTS, R. S., BUSBY, E. C., KENNEDY, A. P., HILL, D. E. & ABRAHAM, R. T. 1998. Inhibition of Phosphoinositide 3-Kinase Related Kinases by the Radiosensitizing Agent Wortmannin. Cancer Research, 58, 4375-4382. SARTORI, A. A., LUKAS, C., COATES, J., MISTRIK, M., FU, S., BARTEK, J., BAER, R., LUKAS, J. & JACKSON, S. P. 2007. Human CtIP promotes DNA end resection. Nature, 450, 509-514. SATO, N., MIZUMOTO, K., NAKAMURA, M. & TANAKA, M. 2000. Radiation-Induced Centrosome Overduplication and Multiple Mitotic Spindles in Human Tumor Cells. Experimental Cell Research, 255, 321-326. SCHMITT, C. A., FRIDMAN, J. S., YANG, M., LEE, S., BARANOV, E., HOFFMAN, R. M. & LOWE, S. W. 2002. A Senescence Program Controlled by p53 and p16INK4a Contributes to the Outcome of Cancer Therapy. Cell, 109, 335-346. SCHNITT, S. J. 2010. Classification and prognosis of invasive breast cancer: from morphology to molecular taxonomy. Mod Pathol, 23, S60-S64. SCHWARZE, S. R., FU, V. X., DESOTELLE, J. A., KENOWSKI, M. L. & JARRARD, D. F. 2005. The Identification of Senescence-Specific Genes during the Induction of Senescence in Prostate Cancer Cells. Neoplasia, 7, 816-823. SERRANO, M., LIN, A. W., MCCURRACH, M. E., BEACH, D. & LOWE, S. W. 1997. Oncogenic ras Provokes Premature Cell Senescence Associated with Accumulation of p53 and p16INK4a. Cell, 88, 593-602. SHAY, J. W. & RONINSON, I. B. 2004. Hallmarks of senescence in carcinogenesis and cancer therapy. Oncogene, 23, 2919-2933. SHERR, C. J. & ROBERTS, J. M. 1995. Inhibitors of mammalian G1 cyclin-dependent kinases. Genes & Development, 9, 1149-1163. SHIEH, S. Y., AHN, J., TAMAI, K., TAYA, Y. & PRIVES, C. 2000. The human homologs of checkpoint kinases Chk1 and Cds1 (Chk2) phosphorylate, p53 at multiple DNA damage-inducible sites. Genes and Development, 14, 289-300. SØRLIE, T., PEROU, C. M., TIBSHIRANI, R., AAS, T., GEISLER, S., JOHNSEN, H., HASTIE, T., EISEN, M. B., VAN DE RIJN, M., JEFFREY, S. S., THORSEN, T., QUIST, H., MATESE, J. C., BROWN, P. O., BOTSTEIN, D., LØNNING, P. E. & BØRRESEN-DALE, A.-L. 2001. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proceedings of the National Academy of Sciences of the United States of America, 98, 10869-10874. STANGEL, D., ERKAN, M., BUCHHOLZ, M., GRESS, T., MICHALSKI, C., RAULEFS, S., FRIESS, H. & KLEEFF, J. 2015. Kif20a inhibition reduces migration and invasion of pancreatic cancer cells. Journal of Surgical Research, 197, 91-100. STAVRIDI, E. S. & HALAZONETIS, T. D. 2005. Nbs1 moving up in the world. Nat Cell Biol, 7, 648-650. STEWART, D. J. 2007. Mechanisms of resistance to cisplatin and carboplatin. Critical Reviews in Oncology/Hematology, 63, 12-31. SULLI, G., DI MICCO, R. & DI FAGAGNA, F. D. A. 2012. Crosstalk between chromatin state and DNA damage response in cellular senescence and cancer. Nat Rev Cancer, 12, 709-720. SUNTERS, A., MADUREIRA, P. A., POMERANZ, K. M., AUBERT, M., BROSENS, J. J., COOK, S. J., BURGERING, B. M. T., COOMBES, R. C. & LAM, E. W.-F. 2006. Paclitaxel-Induced Nuclear Translocation of FOXO3a in Breast Cancer Cells Is Mediated by c-Jun NH2-Terminal Kinase and Akt. Cancer Research, 66, 212-220. SWANTON, C. & JONES, N. 2001. Strategies in subversion: de-regulation of the mammalian cell cycle by viral gene products. International Journal of Experimental Pathology, 82, 3-13. SYMINGTON, L. S. & GAUTIER, J. 2011. Double-Strand Break End Resection and Repair Pathway Choice. Annual Review of Genetics, 45, 247-271.

236

TAN, Y., RAYCHAUDHURI, P. & COSTA, R. H. 2007. Chk2 Mediates Stabilization of the FoxM1 Transcription Factor To Stimulate Expression of DNA Repair Genes. Molecular and Cellular Biology, 27, 1007-1016.

TANIUCHI, K., NAKAGAWA, H., NAKAMURA, T., EGUCHI, H., OHIGASHI, H., ISHIKAWA, O., KATAGIRI, T. & NAKAMURA, Y. 2005. Down-regulation of RAB6KIFL/KIF20A, a Kinesin Involved with Membrane Trafficking of Discs Large Homologue 5, Can Attenuate Growth of Pancreatic Cancer Cell. Cancer Research, 65, 105-112.

TCHERNIUK, S., SKOUFIAS, D. A., LABRIERE, C., RATH, O., GUERITTE, F., GUILLOU, C. & KOZIELSKI, F. 2010. Relocation of Aurora B and Survivin from Centromeres to the Central Spindle Impaired by a Kinesin-Specific MKLP-2 Inhibitor. Angewandte Chemie International Edition, 49, 8228- 8231. TE POELE, R. H., OKOROKOV, A. L., JARDINE, L., CUMMINGS, J. & JOEL, S. P. 2002. DNA Damage Is Able to Induce Senescence in Tumor Cells in Vitro and in Vivo. Cancer Research, 62, 1876- 1883. TEH, M.-T., GEMENETZIDIS, E., CHAPLIN, T., YOUNG, B. D. & PHILPOTT, M. P. 2010. Upregulation of FOXM1 induces genomic instability in human epidermal keratinocytes. Molecular Cancer, 9, 45-45. THORN, C. F., OSHIRO, C., MARSH, S., HERNANDEZ-BOUSSARD, T., MCLEOD, H., KLEIN, T. E. & ALTMAN, R. B. 2011. Doxorubicin pathways: pharmacodynamics and adverse effects. Pharmacogenetics and Genomics, 21, 440-446. TIMOFEEV, O., CIZMECIOGLU, O., SETTELE, F., KEMPF, T. & HOFFMANN, I. 2010. Cdc25 Phosphatases Are Required for Timely Assembly of CDK1-Cyclin B at the G(2)/M Transition. The Journal of Biological Chemistry, 285, 16978-16990. TRAN, H. M., SHI, G., LI, G., CARNEY, J. P., O'MALLEY JR, B. & LI, D. 2004. Mutant Nbs1 enhances cisplatin-induced DNA damage and cytotoxicity in head and neck cancer. Otolaryngology - Head and Neck Surgery, 131, 477-484. TRAVIS, R. C. & KEY, T. J. 2003. Oestrogen exposure and breast cancer risk. Breast Cancer Research : BCR, 5, 239-247. TURNER, N. C. & JONES, A. L. 2008. Management of breast cancer—Part I. UDDIN, S., AHMED, M., HUSSAIN, A., ABUBAKER, J., AL-SANEA, N., ABDULJABBAR, A., ASHARI, L. H., ALHOMOUD, S., AL-DAYEL, F., JEHAN, Z., BAVI, P., SIRAJ, A. K. & AL-KURAYA, K. S. 2011. Genome-Wide Expression Analysis of Middle Eastern Colorectal Cancer Reveals FOXM1 as a Novel Target for Cancer Therapy. The American Journal of Pathology, 178, 537-547. UZIEL, T., LERENTHAL, Y., MOYAL, L., ANDEGEKO, Y., MITTELMAN, L. & SHILOH, Y. 2003. Requirement of the MRN complex for ATM activation by DNA damage. The EMBO Journal, 22, 5612-5621. VALE, R. D., FUNATSU, T., PIERCE, D. W., ROMBERG, L., HARADA, Y. & YANAGIDA, T. 1996. Direct observation of single kinesin molecules moving along microtubules. Nature, 380, 451-453. VALE, R. D., REESE, T. S. & SHEETZ, M. P. 1985. Identification of a novel force-generating protein, kinesin, involved in microtubule-based motility. Cell, 42, 39-50. VAN DER GROEP, P., VAN DER WALL, E. & VAN DIEST, P. J. 2011. Pathology of hereditary breast cancer. Cellular Oncology (Dordrecht), 34, 71-88. VARON, R., VISSINGA, C., PLATZER, M., CEROSALETTI, K. M., CHRZANOWSKA, K. H., SAAR, K., BECKMANN, G., SEEMANOVÁ, E., COOPER, P. R., NOWAK, N. J., STUMM, M., WEEMAES, C. M. R., GATTI, R. A., WILSON, R. K., DIGWEED, M., ROSENTHAL, A., SPERLING, K., CONCANNON, P. & REIS, A. 1998. Nibrin, a Novel DNA Double-Strand Break Repair Protein, Is Mutated in Nijmegen Breakage Syndrome. Cell, 93, 467-476. VERMEULEN, K., VAN BOCKSTAELE, D. R. & BERNEMAN, Z. N. 2003. The cell cycle: a review of regulation, deregulation and therapeutic targets in cancer. Cell Proliferation, 36, 131-149.

237

VERWEIJ, J., CLAVEL, M. & CHEVALIER, B. 1994. Paclitaxel (Taxol) and docetaxel (Taxotere): not simply two of a kind. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO, 5, 495-505. VICENTE, J. J. & WORDEMAN, L. Mitosis, microtubule dynamics and the evolution of kinesins. Experimental Cell Research. VICHAI, V. & KIRTIKARA, K. 2006. Sulforhodamine B colorimetric assay for cytotoxicity screening. Nat. Protocols, 1, 1112-1116. VILENCHIK, M. M. & KNUDSON, A. G. 2003. Endogenous DNA double-strand breaks: Production, fidelity of repair, and induction of cancer. Proceedings of the National Academy of Sciences of the United States of America, 100, 12871-12876. VITALE, I., GALLUZZI, L., CASTEDO, M. & KROEMER, G. 2011. Mitotic catastrophe: a mechanism for avoiding genomic instability. Nat Rev Mol Cell Biol, 12, 385-392. VOGEL, C. L., COBLEIGH, M. A., TRIPATHY, D., GUTHEIL, J. C., HARRIS, L. N., FEHRENBACHER, L., SLAMON, D. J., MURPHY, M., NOVOTNY, W. F., BURCHMORE, M., SHAK, S., STEWART, S. J. & PRESS, M. 2002. Efficacy and Safety of Trastuzumab as a Single Agent in First-Line Treatment of HER2-Overexpressing Metastatic Breast Cancer. Journal of Clinical Oncology, 20, 719-726. VON HOFF, D. D., LAYARD, M. W., BASA, P., DAVIS, J. H. L., VON HOFF, A. L., ROZENCWEIG, M. & MUGGIA, F. M. 1979. Risk Factors for Doxorubicin-lnduced Congestive Heart Failure. Annals of Internal Medicine, 91, 710-717. WALCZAK, C. E. & MITCHISON, T. J. 1996. Kinesin-Related Proteins at Mitotic Spindle Poles: Function and Regulation. Cell, 85, 943-946. WALL, M., POORTINGA, G., STANLEY, K. L., LINDEMANN, R. K., BOTS, M., CHAN, C. J., BYWATER, M. J., KINROSS, K. M., ASTLE, M. V., WALDECK, K., HANNAN, K. M., SHORTT, J., SMYTH, M. J., LOWE, S. W., HANNAN, R. D., PEARSON, R. B., JOHNSTONE, R. W. & MCARTHUR, G. A. 2013. The mTORC1 inhibitor everolimus prevents and treats Eμ-Myc lymphoma by restoring WANG, C.-M., LIU, R., WANG, L., NASCIMENTO, L., BRENNAN, V. C. & YANG, W.-H. 2014. SUMOylation of FOXM1B Alters Its Transcriptional Activity on Regulation of MiR-200 Family and JNK1 in MCF7 Human Breast Cancer Cells. International Journal of Molecular Sciences, 15, 10233-10251. WAN, L.-Y., DENG, J., XIANG, X.-J., ZHANG, L., YU, F., CHEN, J., SUN, Z., FENG, M. & XIONG, J.-P. 2015. miR-320 enhances the sensitivity of human colon cancer cells to chemoradiotherapy in vitro by targeting FOXM1. Biochemical and Biophysical Research Communications, 457, 125-132. WANG, I.-C., CHEN, Y.-J., HUGHES, D., PETROVIC, V., MAJOR, M. L., PARK, H. J., TAN, Y., ACKERSON, T. & COSTA, R. H. 2005a. Forkhead Box M1 Regulates the Transcriptional Network of Genes Essential for Mitotic Progression and Genes Encoding the SCF (Skp2-Cks1) Ubiquitin Ligase. Molecular and Cellular Biology, 25, 10875-10894. WANG, I.-C., CHEN, Y.-J., HUGHES, D. E., ACKERSON, T., MAJOR, M. L., KALINICHENKO, V. V., COSTA, R. H., RAYCHAUDHURI, P., TYNER, A. L. & LAU, L. F. 2008. FoxM1 Regulates Transcription of JNK1 to Promote the G1/S Transition and Tumor Cell Invasiveness. Journal of Biological Chemistry, 283, 20770-20778. WANG, X., KIYOKAWA, H., DENNEWITZ, M. B. & COSTA, R. H. 2002. The Forkhead Box m1b transcription factor is essential for hepatocyte DNA replication and mitosis during mouse liver regeneration. Proceedings of the National Academy of Sciences, 99, 16881-16886. WANG, Z., AHMAD, A., LI, Y., BANERJEE, S., KONG, D. & SARKAR, F. H. 2010. Forkhead box M1 transcription factor: A novel target for cancer therapy. Cancer Treatment Reviews, 36, 151- 156. WANG, Z., BANERJEE, S., KONG, D., LI, Y. & SARKAR, F. H. 2007. Down-regulation of Forkhead Box M1 Transcription Factor Leads to the Inhibition of Invasion and Angiogenesis of Pancreatic Cancer Cells. Cancer Research, 67, 8293-8300.

238

WANG, Z., GOULET III, R., STANTON, K. J., SADARIA, M. & NAKSHATRI, H. 2005b. Differential effect of anti-apoptotic genes Bcl-xL and c-FLIP on sensitivity of MCF-7 breast cancer cells to paclitaxel and docetaxel. Anticancer Research, 25, 2367-2379. WANI, M. C., TAYLOR, H. L., WALL, M. E., COGGON, P. & MCPHAIL, A. T. 1971. Plant antitumor agents. VI. The isolation and structure of taxol, a novel antileukemic and antitumor agent from Taxus brevifolia [12]. Journal of the American Chemical Society, 93, 2325-2327. WEAVER, B. A. 2014. How Taxol/paclitaxel kills cancer cells. Molecular Biology of the Cell, 25, 2677- 2681. WEAVER, B. A. A. & CLEVELAND, D. W. 2005. Decoding the links between mitosis, cancer, and chemotherapy: The mitotic checkpoint, adaptation, and cell death. Cancer Cell, 8, 7-12. WEBER, A. M. & RYAN, A. J. 2015. ATM and ATR as therapeutic targets in cancer. Pharmacology & Therapeutics, 149, 124-138. WELLS, N. J., WATANABE, N., TOKUSUMI, T., JIANG, W., VERDECIA, M. A. & HUNTER, T. 1999. The C- terminal domain of the Cdc2 inhibitory kinase Myt1 interacts with Cdc2 complexes and is required for inhibition of G(2)/M progression. Journal of Cell Science, 112, 3361-3371. WHITFIELD, M. L., SHERLOCK, G., SALDANHA, A. J., MURRAY, J. I., BALL, C. A., ALEXANDER, K. E., MATESE, J. C., PEROU, C. M., HURT, M. M., BROWN, P. O. & BOTSTEIN, D. 2002. Identification of Genes Periodically Expressed in the Human Cell Cycle and Their Expression in Tumors. Molecular Biology of the Cell, 13, 1977-2000. WIERSTRA, I. & ALVES, J. 2006. FOXM1c is activated by cyclin E/Cdk2, cyclin A/Cdk2, and cyclin A/Cdk1, but repressed by GSK-3α. Biochemical and Biophysical Research Communications, 348, 99-108. WIND, N. S. & HOLEN, I. 2011. Multidrug Resistance in Breast Cancer: From In Vitro Models to Clinical Studies. International Journal of Breast Cancer, 2011. WOLANIN, K., MAGALSKA, A., KUSIO-KOBIALKA, M., PODSZYWALOW-BARTNICKA, P., VEJDA, S., MCKENNA, S. L., MOSIENIAK, G., SIKORA, E. & PIWOCKA, K. 2010. Expression of Oncogenic Kinase Bcr-Abl Impairs Mitotic Checkpoint and Promotes Aberrant Divisions and Resistance to Microtubule-Targeting Agents. Molecular Cancer Therapeutics, 9, 1328-1338. WONSEY, D. R. & FOLLETTIE, M. T. 2005. Loss of the forkhead transcription factor FoxM1 causes centrosome amplification and mitotic catastrophe. Cancer research, 65, 5181-9. WOOSTER, R., BIGNELL, G., LANCASTER, J., SWIFT, S., SEAL, S., MANGION, J., COLLINS, N., GREGORY, S., GUMBS, C., MICKLEM, G., BARFOOT, R., HAMOUDI, R., PATEL, S., RICES, C., BIGGS, P., HASHIM, Y., SMITH, A., CONNOR, F., ARASON, A., GUDMUNDSSON, J., FICENEC, D., KELSELL, D., FORD, T. D., TIMOTHY BISHOP, D., SPURR, N. K., PONDER, B. A. J., EELES, R., PETO, J., DEVILEE, P., CORNELISSE, C., LYNCH, H., NAROD, S., LENOIR, G., EGILSSON, V., BJORK BARKADOTTIR, R., EASTON, D. F., BENTLEY, D. R., FUTREAL, P. A., ASHWORTH, A. & STRATTON, M. R. 1995. Identification of the breast cancer susceptibility gene BRCA2. Nature, 378, 789-792. WU, J., ZHANG, X., ZHANG, L., WU, C.-Y., REZAEIAN, ABDOL H., CHAN, C.-H., LI, J.-M., WANG, J., GAO, Y., HAN, F., JEONG, YUN S., YUAN, X., KHANNA, KUM K., JIN, J., ZENG, Y.-X. & LIN, H.-K. 2012. Skp2 E3 Ligase Integrates ATM Activation and Homologous Recombination Repair by Ubiquitinating NBS1. Molecular Cell, 46, 351-361. XUE, J., LIN, X., CHIU, W.-T., CHEN, Y.-H., YU, G., LIU, M., FENG, X.-H., SAWAYA, R., MEDEMA, R. H., HUNG, M.-C. & HUANG, S. 2014. Sustained activation of SMAD3/SMAD4 by FOXM1 promotes TGF-β–dependent cancer metastasis. The Journal of Clinical Investigation, 124, 564-579. XUE, W., ZENDER, L., MIETHING, C., DICKINS, R. A., HERNANDO, E., KRIZHANOVSKY, V., CORDON- CARDO, C. & LOWE, S. W. 2007. Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas. Nature, 445, 656-660.

239

YAN, G.-R., ZOU, F.-Y., DANG, B.-L., ZHANG, Y., YU, G., LIU, X. & HE, Q.-Y. 2012. Genistein-induced mitotic arrest of gastric cancer cells by downregulating KIF20A, a proteomics study. PROTEOMICS, 12, 2391-2399. YANG, D. K., SON, C. H., LEE, S. K., CHOI, P. J., LEE, K. E. & ROH, M. S. 2009. Forkhead box M1 expression in pulmonary squamous cell carcinoma: correlation with clinicopathologic features and its prognostic significance. Human Pathology, 40, 464-470. YANG, M.-H., CHIANG, W.-C., CHANG, S.-Y., CHEN, P.-M. & WU, K.-J. 2005. Increased NBS1 expression is a poor prognostic marker of head and neck cancer and overexpression of NBS1 promotes transformation through the activation of Akt. Cancer Research, 65, 395. YANG, M.-H., CHIANG, W.-C., CHOU, T.-Y., CHANG, S.-Y., CHEN, P.-M., TENG, S.-C. & WU, K.-J. 2006. Increased NBS1 Expression Is a Marker of Aggressive Head and Neck Cancer and Overexpression of NBS1 Contributes to Transformation. Clinical Cancer Research, 12, 507- 515. YANG, M.-Y., LIN, P.-M., LIU, Y.-C., HSIAO, H.-H., YANG, W.-C., HSU, J.-F., HSU, C.-M. & LIN, S.-F. 2012. Induction of Cellular Senescence by Doxorubicin Is Associated with Upregulated miR- 375 and Induction of Autophagy in K562 Cells. PLoS One, 7, e37205. YAO, K.-M., SHA, M., LU, Z. & WONG, G. G. 1997. Molecular Analysis of a Novel Winged Helix Protein, WIN: EXPRESSION PATTERN, DNA BINDING PROPERTY, AND ALTERNATIVE SPLICING WITHIN THE DNA BINDING DOMAIN. Journal of Biological Chemistry, 272, 19827-19836. YE, H., KELLY, T. F., SAMADANI, U., LIM, L., RUBIO, S., OVERDIER, D. G., ROEBUCK, K. A. & COSTA, R. H. 1997. Hepatocyte nuclear factor 3/fork head homolog 11 is expressed in proliferating epithelial and mesenchymal cells of embryonic and adult tissues. Molecular and Cellular Biology, 17, 1626-1641. YEO, B., TURNER, N. C. & JONES, A. 2014. An update on the medical management of breast cancer. BMJ, 348. YOU, Z., CHAHWAN, C., BAILIS, J., HUNTER, T. & RUSSELL, P. 2005. ATM Activation and Its Recruitment to Damaged DNA Require Binding to the C Terminus of Nbs1. Molecular and Cellular Biology, 25, 5363-5379. YU, J., DESHMUKH, H., PAYTON, J. E., DUNHAM, C., SCHEITHAUER, B. W., TIHAN, T., PRAYSON, R. A., GUHA, A., BRIDGE, J. A., FERNER, R. E., LINDBERG, G. M., GUTMANN, R. J., EMNETT, R. J., SALAVAGGIONE, L., GUTMANN, D. H., NAGARAJAN, R., WATSON, M. A. & PERRY, A. 2011. Array-Based Comparative Genomic Hybridization Identifies CDK4 and FOXM1 Alterations as Independent Predictors of Survival in Malignant Peripheral Nerve Sheath Tumor. Clinical Cancer Research, 17, 1924-1934. YUAN, J. & CHEN, J. 2010. MRE11-RAD50-NBS1 Complex Dictates DNA Repair Independent of H2AX. Journal of Biological Chemistry, 285, 1097-1104. YUAN, Z., ZHANG, X., SENGUPTA, N., LANE, W. S. & SETO, E. 2007. SIRT1 Regulates the Function of the Nijmegen Breakage Syndrome Protein. Molecular Cell, 27, 149-162. ZASADIL, L. M., ANDERSEN, K. A., YEUM, D., ROCQUE, G. B., WILKE, L. G., TEVAARWERK, A. J., RAINES, R. T., BURKARD, M. E. & WEAVER, B. A. 2014. Cytotoxicity of paclitaxel in breast cancer is due to chromosome missegregation on multipolar spindles. Science translational medicine, 6, 229ra43-229ra43. ZENG, J., WANG, L., LI, Q., LI, W., BJÖRKHOLM, M., JIA, J. & XU, D. 2009. FoxM1 is up-regulated in gastric cancer and its inhibition leads to cellular senescence, partially dependent on p27kip1. The Journal of Pathology, 218, 419-427. ZHANG, M. H., MAN, H. T., ZHAO, X. D., DONG, N. I. & MA, S. L. 2014a. Estrogen receptor-positive breast cancer molecular signatures and therapeutic potentials (Review). Biomedical Reports, 2, 41-52. ZHANG, N., WU, X., YANG, L., XIAO, F., ZHANG, H., ZHOU, A., HUANG, Z. & HUANG, S. 2012. FoxM1 Inhibition Sensitizes Resistant Glioblastoma Cells to Temozolomide by Downregulating the Expression of DNA-Repair Gene Rad51. Clinical Cancer Research, 18, 5961-5971.

240

ZHANG, Y., LIU, J., PENG, X., ZHU, C.-C., HAN, J., LUO, J. & RUI, R. 2014b. KIF20A Regulates Porcine Oocyte Maturation and Early Embryo Development. PLoS ONE, 9, e102898. ZHANG, Y. & YANG, J.-M. 2011a. The impact of cellular senescence in cancer therapy: is it true or not? Acta Pharmacologica Sinica, 32, 1199-1207. ZHANG, Y. & YANG, J.-M. 2011b. The impact of cellular senescence in cancer therapy: is it true or not? Acta Pharmacol Sin, 32, 1199-1207. ZHANG, Y., ZHANG, N., DAI, B., LIU, M., SAWAYA, R., XIE, K. & HUANG, S. 2008. FoxM1B Transcriptionally Regulates Vascular Endothelial Growth Factor Expression and Promotes the Angiogenesis and Growth of Glioma Cells. Cancer Research, 68, 8733-8742. ZHAO, F., SIU, M. K. Y., JIANG, L., TAM, K. F., NGAN, H. Y. S., LE, X. F., WONG, O. G. W., WONG, E. S. Y., GOMES, A. R., BELLA, L., KHONGKOW, P., LAM, E. W. F. & CHEUNG, A. N. Y. 2014. Overexpression of Forkhead Box Protein M1 (FOXM1) in Ovarian Cancer Correlates with Poor Patient Survival and Contributes to Paclitaxel Resistance. PLoS One, 9, e113478. ZHAO, S., WENG, Y.-C., YUAN, S.-S. F., LIN, Y.-T., HSU, H.-C., LIN, S.-C. J., GERBINO, E., SONG, M.-H., ZDZIENICKA, M. Z., GATTI, R. A., SHAY, J. W., ZIV, Y., SHILOH, Y. & LEE, E. Y. H. P. 2000. Functional link between ataxia-telangiectasia and Nijmegen breakage syndrome gene products. Nature, 405, 473-477. ZHOU, J., WANG, Y., WANG, Y., YIN, X., HE, Y., CHEN, L., WANG, W., LIU, T. & DI, W. 2014. FOXM1 Modulates Cisplatin Sensitivity by Regulating EXO1 in Ovarian Cancer. PLoS One, 9, e96989. ZHU, C., ZHAO, J., BIBIKOVA, M., LEVERSON, J. D., BOSSY-WETZEL, E., FAN, J.-B., ABRAHAM, R. T. & JIANG, W. 2005. Functional Analysis of Human Microtubule-based Motor Proteins, the Kinesins and Dyneins, in Mitosis/Cytokinesis Using RNA Interference. Molecular Biology of the Cell, 16, 3187-3199. ZONA, S., BELLA, L., BURTON, M. J., NESTAL DE MORAES, G. & LAM, E. W. F. 2014. FOXM1: An emerging master regulator of DNA damage response and genotoxic agent resistance. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms, 1839, 1316-1322. ZOU, L. & ELLEDGE, S. J. 2003. Sensing DNA Damage Through ATRIP Recognition of RPA-ssDNA Complexes. Science, 300, 1542-1548.

SUZUKI, N., HAZAMA, S., UENO, T., MATSUI, H., SHINDO, Y., IIDA, M., YOSHIMURA, K., YOSHINO, S., TAKEDA, K. & OKA, M. 2014. A Phase I Clinical Trial of Vaccination With KIF20A-derived Peptide in Combination With Gemcitabine For Patients With Advanced Pancreatic Cancer. Journal of Immunotherapy (Hagerstown, Md. : 1997), 37, 36-42.

241